Ambulatory gynaecology: guidelines and economic analysis

AMBULATORY GYNAECOLOGY: GUIDELINES AND

ECONOMIC ANALYSIS

Natalie Ann MacKinnon Cooper

A thesis submitted to the University of Birmingham for the degree of

DOCTOR OF PHILOSOPHY

School of Clinical and Experimental Medicine

College of Medical and Dental Sciences

University of Birmingham

June 2013

University of Birmingham Research Archive

e-theses repository This unpublished thesis/dissertation is copyright of the author and/or third parties. The intellectual property rights of the author or third parties in respect of this work are as defined by The Copyright Designs and Patents Act 1988 or as modified by any successor legislation. Any use made of information contained in this thesis/dissertation must be in accordance with that legislation and must be properly acknowledged. Further distribution or reproduction in any format is prohibited without the permission of the copyright holder.

ABSTRACT

The aim of this thesis was to investigate the role of outpatient hysteroscopy in modern

gynaecological care by conducting a series of systematic reviews and meta-analyses to

examine how the procedure can be optimised to reduce pain and by performing a cost

effectiveness analysis.

The systematic reviews concluded that women undergoing outpatient hysteroscopy should

take simple analgesia beforehand and that the hysteroscopist should adopt a vaginoscopic

approach using a small diameter, rigid hysteroscope and normal saline as the distension

medium. If dilatation of the cervix is required this should be done under a paracervical block.

These findings were incorporated into a clinical guideline and the quality of the evidence that

the reviews provided was assessed using the SIGN and GRADE methods. A comparison of

the assessments found that they gave varying estimates of the quality of evidence and that

neither offered a perfect solution to the assessment of evidence quality when writing clinical

guidance.

The economic analysis found that initial testing with outpatient hysteroscopy was the most

cost-effective testing strategy for investigation of heavy menstrual bleeding when compared

to other diagnostic tests, regardless of a woman’s wish for future fertility or prior treatment

with a levonorgestrel intrauterine system.

DEDICATION

To my family

ACKNOWLEDGEMENTS

The work for this thesis was conducted between February 2009 and June 2013 whilst I was

working at Birmingham Women’s Hospital, initially as a Research Fellow (funded by the

Outpatient Polyp Treatment (OPT) Trial, which was funded by the National Institute for

Health Research, Health Technology Assessment Programme) and then as a clinical trainee in

obstetrics and gynaecology. There are many people who have helped and supported me and I

would like to acknowledge them for this.

Thank you to Paul Smith for acting as a second reviewer for the systematic reviews. Thank

you to Lee Middleton and Jon Deeks for their advice and help with statistical analysis. Thank

you to Chris Kremer who was a co-author on the ‘Best Practice in Outpatient Hysteroscopy’

guideline.

Pelham Barton provided invaluable support for the cost-effectiveness analysis by advising on

the design of the decision model and managing the economic analyses.

Thank you to the following gynaecologists for acting as my ‘expert panel’ and responding to

my surveys of opinion regarding the decision model and clinical management scenarios:

Dr Mary Connor, Royal Hallamshire Hospital, Sheffield, UK

Prof. Sian Jones, Bradford Teaching Hospitals Foundation Trust, Bradford, UK

Mr. Tyrone Carpenter, Poole Hospital NHS Foundation Trust, Poole, UK

Mr Matthew Parsons, Birmingham Women’s Hospital, Birmingham, UK

Ms Mamta Patak, Worcestershire Royal Hospital, Worcester, UK

Mr Kevin Phillips, Castle Hill Hospital, Hull, UK

Dr Hanny Pijnenborg, TweeSteden Ziekenhuis, Tilburg, NL

Mr Nick Raine-Fenning, University of Nottingham, Nottingham, UK

Mr Robert Richardson, Chelsea and Westminster Hospital, London, UK

Dr Anthony Roberts, Queens Hospital, Burton-upon-Trent, UK

Ms Manjeet Shehmar, Birmingham Women’s Hospital, Birmingham, UK

Dr Andreas Thurkow, Saint Lucas Andreas Ziekenhuis, Amsterdam, NL

Dr Sebastiaan Veersema, St Antonius Hospital , Utrecht, NL

Dr. Lucet van der Voet, Deventer Ziekenhuis, NL

Dr. Maria Breijer, Amsterdam Medical Centre, Amsterdam, NL

Prof. Ben Willem Mol, University of Amsterdam, Amsterdam NL

Thank you to Dr. Raji Ganesan for providing advice regarding plausible false diagnoses for

endometrial biopsy samples and to Dr. Ioannis Gallos and Prof. Janesh Gupta for providing

their follow up data for women with endometrial disease being treated with the levonorgestrel

intrauterine system.

Thank you to my colleagues Siobhan O’Connor, Tracy Bingham, Laura Gennard and Avril

Williams for their cups of tea, ‘pep talks’ and friendship.

Khalid Khan taught me how to use meta-analysis software and provided meta-regression

analysis for the systematic review of local anaesthetic. As a supervisor he taught me

methodology for conducting systematic reviews and helped me to optimise my work for

http://nl.linkedin.com/company/sint-lucas-andreas-ziekenhuis?trk=ppro_cprof

publication. Thank you to Arri Coomarasamy who stepped in as supervisor when Khalid left

the University of Birmingham for the bright lights of London.

My greatest thanks have to go to my lead supervisor Justin Clark for introducing me to

research and for his continual encouragement, advice and support. He has an inexhaustible

enthusiasm for teaching and improving gynaecology care and has given me numerous

opportunities that have allowed me to develop as an academic and a clinician.

Finally I would like to thank my parents, Bruce and Catherine Cooper, my sister, Alexandra

Cooper and my husband Lexi Cherniavsky for their love and encouragement, for allowing me

to spend hours in front of my computer and for being so very proud of me.

CONTENTS

CHAPTER 1 THESIS OVERVIEW AND OBJECTIVES 1

PROJECT OBJECTIVES 2

CHAPTER 2 INTRODUCTION TO A SERIES OF SYSTEMATIC REVIEWS AND

META-ANALYSES WHICH INVESTIGATE PAIN REDUCTION IN OUTPATIENT

HYSTEROSCOPY 3

CHAPTER 3 METHODS FOR THE SEVEN SYSTEMATIC REVIEWS AND META-

ANALYSES WHICH INVESTIGATE PAIN REDUCTION IN OUTPATIENT

HYSTEROSCOPY 9

TOPICS FOR INVESTIGATION 9

DATA SOURCES, SEARCHES AND STUDY SELECTION 10

DATA EXTRACTION 15

DATA SYNTHESIS 24

CHAPTER 4 RESULTS OF THE SEVEN SYSTEMATIC REVIEWS AND META-

ANALYSES WHICH INVESTIGATE PAIN REDUCTION IN OUTPATIENT

HYSTEROSCOPY 26

RESULTS OF THE SYSTEMATIC REVIEW AND META-ANALYSIS OF LOCAL ANAESTHESIA FOR PAIN CONTROL DURING

OUTPATIENT HYSTEROSCOPY 26

RESULTS OF THE SYSTEMATIC REVIEW OF ANALGESIA FOR PAIN CONTROL DURING OUTPATIENT HYSTEROSCOPY 43

RESULTS OF THE SYSTEMATIC REVIEW OF CONSCIOUS SEDATION FOR PAIN CONTROL DURING OUTPATIENT

HYSTEROSCOPY 49

RESULTS OF THE SYSTEMATIC REVIEW OF THE EFFECT OF CERVICAL PREPARATION ON PAIN DURING OUTPATIENT

HYSTEROSCOPY 51

RESULTS OF THE SYSTEMATIC REVIEW OF THE EFFECT ON PAIN OF THE VAGINOSCOPIC APPROACH TO OUTPATIENT

HYSTEROSCOPY 59

RESULTS OF THE SYSTEMATIC REVIEW OF THE EFFECT OF DISTENSION MEDIA ON PAIN DURING OUTPATIENT

HYSTEROSCOPY 66

RESULTS OF THE SYSTEMATIC REVIEW OF THE EFFECT ON PAIN OF THE TYPE OF HYSTEROSCOPE USED FOR


CHAPTER 5 DISCUSSION REGARDING THE RESULTS OF THE SYSTEMATIC

REVIEWS AND META-ANALYSES WHICH INVESTIGATE PAIN REDUCTION IN


PRINCIPAL FINDINGS OF THE REVIEWS 79

STRENGTHS AND LIMITATIONS OF THE REVIEWS 82

COMPARISON WITH OTHER STUDIES 85

CLINICAL IMPLICATIONS OF THE REVIEWS 89

UNANSWERED QUESTIONS AND FUTURE RESEARCH 93

SUMMARY 96

CHAPTER 6 EVIDENCE QUALITY IN CLINICAL GUIDELINES: A COMPARISON OF

TWO METHODS 97

METHODS FOR COMPARING SIGN AND GRADE FOR ASSESSING QUALITY OF EVIDENCE 98

RESULTS OF THE COMPARISON OF SIGN AND GRADE FOR ASSESSING QUALITY OF EVIDENCE 104

DISCUSSION OF THE COMPARISON OF SIGN AND GRADE FOR ASSESSING QUALITY OF EVIDENCE 106

CHAPTER 7 INTRODUCTION TO THE ECONOMIC ANALYSIS OF DIAGNOSTIC

STRATEGIES FOR THE INVESTIGATION OF HEAVY MENSTRUAL BLEEDING 111

HEAVY MENSTRUAL BLEEDING 111

CAUSES OF HEAVY MENSTRUAL BLEEDING 112

DIAGNOSIS OF HEAVY MENSTRUAL BLEEDING 113

CURRENT DIAGNOSTIC PATHWAYS FOR HEAVY MENSTRUAL BLEEDING 115

LITERATURE REVIEW OF COST-EFFECTIVENESS STUDIES FOR THE DIAGNOSTIC WORK UP OF HEAVY MENSTRUAL

BLEEDING 116

CURRENT TREATMENT OF HEAVY MENSTRUAL BLEEDING 119

DEFINING TREATMENT SUCCESS IN HEAVY MENSTRUAL BLEEDING 122

CHAPTER 8 METHODS FOR THE ECONOMIC ANALYSIS OF DIAGNOSTIC


CONSTRUCTION OF THE DECISION MODEL 124

ADAPTATIONS OF THE BASE CASE TREE TO ASSESS ALTERNATIVE CLINICAL SCENARIOS 139

CLINICAL DATA COLLECTION 142

RESULTS OF CLINICAL DATA COLLECTION 148

COSTS 175

METHODS FOR THE COST-EFFECTIVENESS ANALYSIS 180

CHAPTER 9 RESULTS OF THE ECONOMIC ANALYSIS OF DIAGNOSTIC


DETERMINISTIC RESULTS: BASE CASE 188

PROBABILISTIC SENSITIVITY ANALYSIS RESULTS: BASE CASE 194

DETERMINISTIC SENSITIVITY ANALYSIS RESULTS 203

DETERMINISTIC RESULTS: WOMEN BEING MANAGED DURING MULTIPLE CLINIC APPOINTMENTS 206

PROBABILISTIC SENSITIVITY ANALYSIS RESULTS: WOMEN BEING MANAGED DURING MULTIPLE CLINIC

APPOINTMENTS 209

DETERMINISTIC RESULTS: PRIOR TREATMENT WITH THE LNG-IUS 219

PROBABILISTIC SENSITIVITY ANALYSIS RESULTS: PRIOR TREATMENT WITH THE LNG-IUS 224

DETERMINISTIC ANALYSIS: WOMEN WISHING TO PRESERVE THEIR FERTILITY 229

PROBABILISTIC SENSITIVITY ANALYSIS RESULTS: WOMEN WISHING TO PRESERVE THEIR FERTILITY 233

CHAPTER 10 DISCUSSION OF THE ECONOMIC ANALYSIS OF DIAGNOSTIC


DISCUSSION OF THE MAIN FINDINGS 240

STRENGTHS AND LIMITATIONS 248

COMPARISONS WITH EXISTING GUIDANCE 257

CHAPTER 11 CONCLUSIONS 258

SUMMARY OF THE FINDINGS OF THE SYSTEMATIC REVIEWS OF OUTPATIENT HYSTEROSCOPY 258

SUMMARY OF THE COMPARISON OF THE SIGN AND GRADE METHODS FOR ASSESSING EVIDENCE QUALITY 260

SUMMARY OF FINDINGS OF THE ECONOMIC ANALYSIS OF HMB 260

SUMMARY OF THE ROLE OF OUTPATIENT HYSTEROSCOPY AS INVESTIGATED BY THIS THESIS 262

APPENDIX 1 SEARCH STRATEGIES FOR THE SYSTEMATIC REVIEWS OF


APPENDIX 2 DATA RETRIEVAL FORM FOR SYSTEMATIC REVIEW AND META-

ANALYSIS OF LOCAL ANAESTHESIA FOR PAIN CONTROL DURING OUTPATIENT

HYSTEROSCOPY 277

APPENDIX 3 JADAD METHOD FOR SCORING THE QUALITY OF RANDOMISED

CONTROLLED TRIALS 282

APPENDIX 4 ‘BEST PRACTICE IN OUTPATIENT HYSTEROSCOPY’ GUIDELINE 284

APPENDIX 5 GRADE TABLES FOR ‘BEST PRACTICE IN HYSTEROSCOPY’

GUIDELINE RECOMMENDATIONS WITH CORRESPONDING SIGN GRADINGS 305

APPENDIX 6 DECISION TREES FOR THE ECONOMIC ANALYSIS OF HEAVY

MENSTRUAL BLEEDING 318

APPENDIX 7 SEARCH STRATEGIES FOR COLLECTION OF DATA TO POPULATE

THE DECISION TREES FOR THE ECONOMIC ANALYSIS OF HEAVY MENSTRUAL

BLEEDING 351

REFERENCES 371

LIST OF FIGURES

Chapter 2

Figure 2.1 Diagram showing the different methods of local anaesthetic

administration for outpatient hysteroscopy

4

Chapter 4

Figure 4.1 Study selection process for systematic review of local anaesthetic for

pain relief during outpatient hysteroscopy

27

Figure 4.2 Jadad Quality Assessment of Studies Using Local Anaesthetic for

Out-patient Hysteroscopy

36

Figure 4.3 Forest plot showing the results of meta-analysis of studies that

examine the use of local anaesthetic for reducing pain during out-patient

hysteroscopy. Results overall and sub-grouped according to method of

administration and quality

38


examine the use of intracervical injection without the study whose data were

transformed

39


examine the use anaesthetics sub-grouped into injectable and topical application

41

Figure 4.6 Forest plot showing the results of meta-analysis examining the

incidence of vasovagal episodes in studies examining the use of local anaesthetic

for out-patient hysteroscopy

42

Figure 4.7 Study selection process for the systematic reviews of the effect of

analgesia on pain during outpatient hysteroscopy

43

Figure 4.8. Methodological quality assessment of studies using analgesia for

outpatient hysteroscopy

46

Figure 4.9 Study selection process for the systematic reviews of the effect of

conscious sedation on pain during outpatient hysteroscopy

49

Figure 4.10 Study selection process for the systematic review of the effect of

cervical preparation on pain during outpatient hysteroscopy

51

Figure 4.11 Forest plots showing the results of meta-analysis of adverse effects

and failed procedures when prostaglandins are given prior to outpatient

hysteroscopy

58

Figure 4.12 Study selection process for systematic review of the effect on pain of

the vaginoscopic approach to outpatient hysteroscopy

59

Figure 4.13 Methodological Quality Assessment of Studies Examining the effect

of the vaginoscopic approach on pain during outpatient hysteroscopy

60


examine the use of a vaginoscopic approach to outpatient hysteroscopy

64

Figure 4.15 Forest plot showing the results of meta-analysis of the high quality

studies that examine the use of a vaginoscopic approach to outpatient

hysteroscopy

65

Figure 4.16 Forest plot showing the results of meta-analysis of the studies that

report feasibility of the vaginoscopic approach to outpatient hysteroscopy

65

Figure 4.17 Study selection process for systematic review of the effect of

distension media on pain during outpatient hysteroscopy

67

Figure 4.18 Forest plot showing the effect of distension media on procedural

pain during outpatient hysteroscopy

71

Figure 4.19 Forest plot showing the effect of distension media on shoulder tip

pain during hysteroscopy

72

Figure 4.20 Forest plot showing the effect of distension media on vasovagal

episodes during outpatient hysteroscopy

73

Figure 4.21 Forest plot showing the effect of distension media on procedural

time for outpatient hysteroscopy

73

Figure 4.22 Study selection process for systematic review of the effect on pain of

the type of hysteroscope used for outpatient hysteroscopy

75

Chapter 6

Figure 6.1. Quality rating according to SIGN and GRADE methodology and the

difference between the two scores.

104

Figure 6.2. Reasons for downgrading the quality of evidence for guideline

recommendations when using the GRADE system.

105

Chapter 8

Figure 8.1 Example decision tree for evaluating the cost-effectiveness of

diagnostic testing in heavy menstrual bleeding

127

Chapter 9

Figure 9.1 Cost-effectiveness plane showing the results of deterministic analysis

of the strategies for investigation of women with heavy menstrual bleeding

192

Figure 9.2 Cost-effectiveness plane showing the results of deterministic analysis

of the strategies for investigation of women with heavy menstrual bleeding-

hysterectomy excluded

193

Figure 9.3 Scatterplot showing the uncertainty in costs and effectiveness within

the model for each of the individual strategies for investigation of women with

heavy menstrual bleeding

194

Figure 9.4 Cost-effectiveness acceptability frontier showing the optimal

investigative strategy across a range of willingness to pay thresholds for the

sensitivity analysis of strategies to investigate women with heavy menstrual

bleeding

195

Figure 9.5 Cost-effectiveness plane (a) and cost-effectiveness acceptability curve

(b): Outpatient hysteroscopy alone strategy relative to the LNG-IUS alone

strategy

197


(b): Outpatient hysteroscopy and endometrial biopsy strategy relative to the

outpatient hysteroscopy alone strategy

198


(b): Transvaginal ultrasound scan alone strategy relative to the LNG-IUS alone

strategy

199


(b): Saline infusion sonography alone strategy relative to the LNG-IUS alone

strategy

200


(b): Outpatient hysteroscopy alone strategy relative to the transvaginal

ultrasound scan alone strategy

201

Figure 9.10 Cost-effectiveness plane (a) and cost-effectiveness acceptability

curve (b): Outpatient hysteroscopy alone strategy relative to the saline infusion

sonography alone strategy

202

Figure 9.11 Incremental cost-effectiveness ratios between outpatient

hysteroscopy and saline infusion sonography when the cost of saline infusion

sonography is varied

204

Figure 9.12 Incremental cost-effectiveness ratios between non-dominated options

at a variety of prevalence values for polyps / submucous fibroids and

dysfunctional uterine bleeding

205

Figure 9.13 Cost-effectiveness plane showing the results of deterministic

analysis for strategies to investigate women presenting with heavy menstrual

bleeding managed over multiple clinic appointments. (Hysterectomy removed).

208


the model for each of the individual strategies for investigating women with

heavy menstrual bleeding managed over multiple clinic appointments.

209


investigative strategies for women with heavy menstrual bleeding managed over

multiple clinic appointments.

210


curve (b): transvaginal ultrasound scan alone strategy relative to the LNG-IUS

alone strategy for women with heavy menstrual bleeding managed over multiple

clinic appointments.

211


curve (b): saline infusion sonography alone strategy relative to the LNG-IUS

alone strategy for women with heavy menstrual bleeding managed over multiple

clinic appointments.

212


curve (b): outpatient hysteroscopy alone strategy relative to the saline infusion

sonography alone strategy for women with heavy menstrual bleeding managed

over multiple clinic appointments.

213


curve (b): outpatient hysteroscopy with endometrial biopsy strategy relative to

the outpatient hysteroscopy alone strategy for women with heavy menstrual

bleeding managed over multiple clinic appointments

214


curve (b): saline infusion sonography alone strategy relative to the transvaginal

scan alone strategy for women with heavy menstrual bleeding managed over


215


curve (b): saline infusion sonography with endometrial biopsy strategy relative to

the saline infusion sonography alone strategy for women with heavy menstrual

bleeding managed over multiple clinic appointments

216



the saline infusion sonography with endometrial biopsy strategy for women with

heavy menstrual bleeding managed over multiple clinic appointments.

217



the saline infusion sonography alone strategy for women with heavy menstrual

bleeding managed over multiple clinic appointments.

218


analysis for strategies to investigate women with heavy menstrual bleeding with

a LNG-IUS in situ

222


analysis for strategies to investigate women with heavy menstrual bleeding with

a LNG-IUS in situ. (Hysterectomy alone and LNG-IUS alone not shown)

223



heavy menstrual bleeding with a LNG-IUS

224


investigative strategies for women with heavy menstrual bleeding with a LNG-

IUS in situ

225


curve (b): Outpatient hysteroscopy alone strategy relative to the LNG-IUS alone

strategy for women with heavy menstrual bleeding with a LNG-IUS in situ

226


curve (b): Transvaginal scan and endometrial biopsy strategy relative to the

outpatient hysteroscopy alone strategy for women with heavy menstrual bleeding

with a LNG-IUS in situ

227


curve (b): Saline infusion scan alone strategy relative to the outpatient

hysteroscopy alone strategy for women with heavy menstrual bleeding with a

LNG-IUS in situ

228

Figure 9.31 Total costs and effectiveness of the alternative strategies for the

diagnostic work up of HMB for women wishing to preserve their fertility.

(Hysterectomy has been excluded).

231



heavy menstrual bleeding who wish to preserve their fertility

233

Figure 9.33 Cost-effectiveness acceptability frontier showing the preferred

diagnostic strategy over a range of WTP thresholds for women who wish to

preserve their fertility

234


curve (b): Saline infusion sonography alone strategy relative to the LNG-IUS

alone strategy for women wishing to preserve their fertility

235


curve (b): Outpatient hysteroscopy alone strategy relative to the saline infusion

sonography alone strategy for women wishing to preserve their fertility

236


curve (b): Saline infusion sonography and outpatient hysteroscopy strategy

relative to the outpatient hysteroscopy alone strategy for women wishing to


237


curve (b): Transvaginal ultrasound scan alone strategy relative to the LNG-IUS


238


curve (b): Saline infusion sonography alone strategy relative to the transvaginal

ultrasound scan alone strategy for women wishing to preserve their fertility

239

LIST OF TABLES

Chapter 3

Table 3.1. Search terms and selection criteria for the seven systematic reviews

and meta-analyses which investigate pain reduction in outpatient hysteroscopy

12

Table 3.2. Methodological quality assessment of the studies included in the

systematic review of use of local anaesthetic for outpatient hysteroscopy

16

Table 3.3 Methodological quality assessment of the studies included in the

systematic review analgesia for pain control during outpatient hysteroscopy

18


systematic review of conscious sedation for pain control during outpatient

hysteroscopy

19


systematic review of the effect of cervical preparation on pain during outpatient

hysteroscopy

19

Table 3.6 Methodological quality assessment of the studies included in the

systematic review of the effect on pain of the vaginoscopic approach to outpatient

hysteroscopy.

21


systematic review of the effect of the distension medium used on pain during

outpatient hysteroscopy.

22


systematic review of the effect on pain of the type of hysteroscope used for


23

Chapter 4

Tables 4.1a-c Characteristics of the studies included in the systematic review of

use of local anaesthetic during outpatient hysteroscopy, subgrouped according to

distension media.

29

Table 4.2. Characteristics of the selected studies included in the systematic

reviews of analgesia for pain control during outpatient hysteroscopy

44

Table 4.3. Methods, interventions and results for the studies included in the

systematic review of the effect of cervical preparation on pain during outpatient

hysteroscopy

53

Table 4.4. Characteristics of the selected studies included in the systematic

review of the effect of the vaginoscopic approach on pain during outpatient

hysteroscopy.

61

Table 4.5 Characteristics of the selected studies included in the systematic review

of the effect of the distension medium used on pain during outpatient

hysteroscopy

68

Table 4.6 Overview of the studies included in the systematic review of the effect

on pain of the type of hysteroscope used for outpatient hysteroscopy including

interventions and results

77

Chapter 6

Table 6.1 SIGN method for classification of evidence for clinical guidelines

99

Table 6.2 GRADE evidence table examining the evidence for using different

types of distension medium for outpatient hysteroscopy

102

Table 6.3 Score allocated to the quality of evidence gradings as determined by

SIGN and GRADE

103

Table 6.4 Scores applied to the downgrading aspects of the GRADE profile

elements

104

Chapter 7

Table 7.1 Causes of heavy menstrual bleeding

113

Table 7.2 Description of currently used tests for the diagnosis of uterine

pathology

114

Chapter 8

Table 8.1 False diagnoses according to testing modality

135

Table 8.2 First and second line treatments for heavy menstrual bleeding

according to underlying diagnosis in women with no desire to retain fertility

139

Table 8.3 Gold standard diagnostic tests for uterine pathology

143

Table 8.4 Quality criteria for disease prevalence studies

143

Table 8.5 Quality criteria for assessing test accuracy studies

145

Table 8.6 Accuracy value calculation using sensitivity and specificity

146

Table 8.7 Estimated prevalence of uterine pathologies in heavy menstrual

bleeding: Base case

151

Table 8.8 Recalculated disease prevalence for HMB in women refractory to

treatment with a LNG-IUS

152

Table 8.9 Diagnostic test success rate data

155

Table 8.10 Test sensitivity for specific pathologies

161

Table 8.11 Manipulation of false positive rates and their use within the decision

tree

164

Table 8.12 Costs for investigation and treatment of women presenting with heavy

menstrual bleeding

178

Table 8.13 Beta distribution parameters for analysis of the base case decision tree

for investigating women with heavy menstrual bleeding

184

Chapter 9

Table 9.1 Determinist analysis results for the investigation of women with heavy

menstrual bleeding

189

Table 9.2 Deterministic results for the non-dominated strategies for investigation

of women with heavy menstrual bleeding

191

Table 9.3 Incremental cost-effectiveness ratio value for saline infusion

sonography when the cost is varied

203

Table 9.4 Deterministic results of cost-effectiveness analysis for women with

heavy menstrual bleeding managed over multiple clinic appointments

206

Table 9.5 Non-dominated strategies from the analysis of women presenting with

heavy menstrual bleeding managed over multiple clinic appointments

207


heavy menstrual bleeding with a LNG-IUS in situ

220


heavy menstrual bleeding with a LNG-IUS in situ

221



229



231

LIST OF ABBREVIATIONS

AUB Abnormal uterine bleeding

BMI Body mass index

BSO Bilateral salpingo-oophorectomy

BWH Birmingham Women’s Hospital

CEAC Cost-effectiveness acceptability curve

CEAF Cost-effectiveness acceptability frontier

CI Confidence interval

D&C Dilatation and curettage

DUB Dysfunctional uterine bleeding

EA Endometrial ablation

EBx Endometrial biopsy

FIGO International Federation of Gynaecology and Obstetrics

FPR False positive rate

GnRH-a Gonadatrophin releasing hormone antagonist

HMB Heavy menstrual bleeding

HRG Health resources group

HRQL Health related quality of life

ICER Incremental cost-effectiveness ratio

IPD Individual patient data

LNG-IUS Levonorgestrel intrauterine system

NICE National institute for health and clinical excellence

NHS National Health Service

NSAIDs Non-steroidal anti-inflammatories

OPH Outpatient hysteroscopy

QALY Quality adjusted life years

RCT Randomised controlled trial

SIS Saline infusion sonography /scan

SR Systematic review

TAH Total abdominal hysterectomy

TPR True positive rate

TVS Transvaginal ultrasound scan

UAE Uterine artery embolisation

WTP Willingness to pay

1

CHAPTER 1

THESIS OVERVIEW AND OBJECTIVES

In recent years the ‘classical’ inpatient hysteroscopy, which allows gynaecologists to visualise the

lining of the womb has evolved into a diagnostic and therapeutic outpatient procedure. The Royal

College of Obstetricians and Gynaecologists (RCOG) has published standards of care stating that

‘outpatient-based diagnostic services should be available in the community and hospital setting,

including operative procedures for carefully selected cases’(1). For this to be possible, outpatient

hysteroscopy (OPH) needs to be acceptable to patients and to be cost-effective when compared to

alternative outpatient diagnostic testing strategies such as transvaginal ultrasound scan (TVS) and

global endometrial biopsy (EBx).

The aim of this thesis was to investigate the role of outpatient hysteroscopy in modern

gynaecological care. There are two themes within the thesis. The first examines how outpatient

hysteroscopy can be optimised using systematic reviews. A clinical guideline was then produced

to recommend ‘best-practice’ for outpatient hysteroscopy. The quality of the evidence behind the

guideline was then graded and examined using two guideline methods.

The second theme within the thesis looks at the cost-effectiveness of outpatient hysteroscopy and

other diagnostic tests in the investigation of heavy menstrual bleeding. The aim of this analysis is

2

to determine the most cost-effective diagnostic strategy for investigating women who present to

secondary care with HMB.

Project objectives

The thesis objectives are as follows:

to examine different aspects of outpatient hysteroscopy and perform systematic reviews to

determine how best they can be adapted to make the procedure as tolerable as possible for

patients.

to use the data from the systematic reviews to write a ‘Green-top’ style guideline and a

compare the evidence gradings produced with those allocated by the GRADE guideline

process.

to examine strategies for investigating HMB to see if outpatient hysteroscopy is a cost-

effective option when compared to other outpatient diagnostic tests.

3

CHAPTER 2

INTRODUCTION TO A SERIES OF SYSTEMATIC REVIEWS

AND META-ANALYSES WHICH INVESTIGATE PAIN

REDUCTION IN OUTPATIENT HYSTEROSCOPY

Diagnostic and operative outpatient hysteroscopy is feasible, safe and acceptable to women (2-5),

however, the outpatient setting can present many challenges, with the conscious patient being less

forgiving of induced discomfort (6). To examine how outpatient hysteroscopy can be optimised

to reduce pain, different aspects of the procedure can be examined and the available alternative

options evaluated. For example, is using normal saline as a distension medium less painful than

using carbon dioxide? What method of anaesthetising the cervix is the least painful? Can the

procedure be performed without passing a vaginal speculum?

The obvious method of controlling pain is to use pharmaceuticals, whether they are anaesthetics

or analgesics. Pain from the body of the uterus is relayed by T12 to L2 nerve roots which are not

accessible vaginally (7) and thus cannot be anaesthetised, however the splanchnic plexus nerves

(S2-S4) (7) which sense pain from the cervix and lower uterus can be anaesthetised by

administering local anaesthetic to the cervix. Vasovagal reactions during outpatient hysteroscopy

are reported to occur in 20% of women in the general population (8). Nerve stimulation during

passage of instruments through the cervical canal has been suggested as the cause of vasovagal

4

episodes. It has been theorised that blocking the nerves with local anaesthetic may reduce the

incidence of attacks (9-11) but there is also a study which has shown conflicting results (12).

Individual studies examining the effect of local anaesthetics are often imprecise and hence

provide contradictory results (13). Local anaesthetic can be applied as an injection directly into

the body of the cervix (intracervical or direct cervical block) or to the surrounding tissues

(paracervical block), or topically to the uterine cavity (transcervical application) or the ectocervix

(topical application), these methods are shown in Figure 2.1.

Figure 2.1 Diagram showing the different methods of local anaesthetic administration for


A recent review (14) examined the use of paracervical injection for cervical dilatation and uterine

interventions in a variety of obstetric and gynaecological procedures, however, there is no

5

comprehensive review evaluating comparative effectiveness of the whole range of local

anaesthetic modalities for specific procedures.

Analgesics block pain receptors and prevent pain signals being relayed to the central nervous

system. They can be administered before, during or after procedures to reduce pain. Opiate and

non-steroidal inflammatory drugs (NSAID’s) have been used to reduce pain during outpatient

hysteroscopy as has conscious sedation (6), however no guidance exists on use of analgesics for

outpatient hysteroscopy, resulting in eclectic practice.

A compromise between general anaesthesia and analgesics might be to use conscious sedation

(i.e. depression of the central nervous system enabling treatment to be carried out, but during

which verbal contact with the patient is maintained’ (15)) which is widely used in endoscopic

procedures of the gastrointestinal system and in dentistry to make unpleasant interventions more

acceptable but it is less commonly employed in outpatient hysteroscopy. The Royal College of

Anaesthetists has issued guidance on the safe use of conscious sedation by other medical

specialties, recommending that a designated trained staff member be responsible for monitoring

the patient throughout the sedation and that resuscitation facilities be available (15). This has

obvious cost implications on staffing, equipment and recovery areas. To justify these costs there

would need to be a large added benefit of using conscious sedation over less compromising

medication.

A final group of drugs which may alleviate discomfort during outpatient hysteroscopy are those

which can dilate the cervix and potentially aid the passage of the hysteroscope into the uterine

6

cavity. Hysteroscopy is traditionally an inpatient procedure that requires dilatation of the cervix

so that large diameter instruments can be passed into the uterine cavity. Blind mechanical

dilatation risks perforation of the uterus, especially where the cervical canal provides more

resistance i.e. postmenopausal or nulliparous women or those who have had previous cervical

surgery or caesarean section (16;17). Studies examining the use of pharmaceutical

(antiprogestogens, prostaglandins) and mechanical (laminaria) dilatation of the cervix prior to

inpatient hysteroscopy under general anaesthesia have produced conflicting results regarding

their effect on dilatation and trauma during the procedure (18-22).

Prostaglandins are commonly used in gynaecology for inducing dilatation of the cervix for

termination of pregnancy and for removing retained products of conception (23). The natural

progression from this has been for clinicians to use prostaglandins in non-pregnant women to

dilate the cervix and ease access to the uterine cavity for transcervical procedures (24). However,

the evidence is unclear as to whether the dilatory effect is apparent in a non-pregnant cervix and

prostaglandins have unpleasant side effects (nausea, vomiting and excessive bleeding) (25).

Similarly antiprogestogens have also been used to soften the cervix (26) as well as dilatory

materials which when inserted into the cervical canal absorb water and cause a mechanical

dilatation (18;27;28). Again these methods have associated side effects .Thus any benefit may

become outweighed by the associated adverse effects. The technological advances that have

resulted in miniaturisation of endoscopic instruments have done away with the need for routine,

painful cervical dilatation. However, a significant minority of women undergoing an outpatient

procedure will require cervical dilatation with the risks of excessive pain and potential trauma to

the genital tract.

7

As well as investigating pharmaceuticals that can play a role in pain there are technical aspects of

the outpatient hysteroscopy procedure that can be adapted to be less painful. The first method

explored is vaginoscopy, also known as the ‘no-touch’ technique which is an alternative method

for performing hysteroscopy without the need for a vaginal speculum to view the cervix (29) or

cervical instrumentation to grasp and steady it (6). The hysteroscope is inserted into the vagina

which is distended with the selected medium, the external cervical os identified and the

hysteroscope steered into the cervical canal so that the hysteroscopy can be performed.

Distension of the vagina to aid identification of the cervix can be facilitated by a Trendelenburg

tilt and manual closure of the labia minora. Individual papers have suggested that use of the

vaginoscopic technique is feasible and may reduce the pain of outpatient hysteroscopy (30-33)

but no clear, collated summaries of evidence exist.

Once the cervix is traversed and the cavity entered a distending medium needs to be instilled to

visualise the uterine cavity. A variety of fluid (normal saline, water, sorbitol, dextran and glycine

(34-37)) and gaseous (carbon dioxide (38;39)) media have been used. As with any hollow

viscous, distension of the uterine cavity causes pain (irritation of T10 –L2 nerve roots).

Moreover, spilling of distension medium into the abdominal cavity can be associated with

phrenic (C3-5) nerve irritation and referred pain to the shoulder tip and cervical manipulation

may cause vagal stimulation resulting in fainting episodes (6). Image quality is an important

consideration and may vary according to medium used.

The final technical aspect of the procedure to explore is the type of hysteroscope used for the

procedure. Flexible endoscopes are used to investigate the gastrointestinal and respiratory tracts

8

as they allow the operator to negotiate the convoluted anatomy. Flexible hysteroscopes are able to

exploit this advantage when negotiating the cervical canal and viewing the cornuae (40). Studies

have shown that outpatient flexible hysteroscopy is well tolerated (41;42) so this technique was

compared to outpatient hysteroscopy using rigid hysteroscopes.

9

CHAPTER 3

METHODS FOR THE SEVEN SYSTEMATIC REVIEWS AND

META-ANALYSES WHICH INVESTIGATE PAIN REDUCTION

IN OUTPATIENT HYSTEROSCOPY

Topics for investigation

Seven different aspects of outpatient hysteroscopy were evaluated:

1. the use of local anaesthesia

2. the use of analgesics

3. the use of conscious sedation

4. the use of cervical preparation

5. the vaginoscopic approach

6. different distension media

7. the type of hysteroscope

All of the systematic reviews were conducted prospectively, devising a protocol based upon

widely documented methods (43;44).

10

Data Sources, Searches and Study Selection

A comprehensive literature search was conducted for each of the seven topics to identify relevant

studies. The databases searched included Medline (from 1950 to September 2008), EMBASE

(from 1980 to September 2008), CINAHL (from 1981 to September 2008) and the Cochrane

library. The search terms used are shown in Table 3.1 and the full search strategies are shown in

Appendix 1. There were no limits or filters placed on the searches to ensure maximal sensitivity.

The reference sections of selected original articles were checked for relevant papers that had not

already been retrieved by the database searches. The contents lists from two specialist journals

(Gynaecological Surgery and The Journal of Minimally Invasive Gynecology) from November

1993 until 2008 were scrutinised for relevant studies.

The titles and abstracts from the electronic literature searches were reviewed in duplicate (see

acknowledgements). The citations were selected if they appeared to fulfil the selection criteria.

The selection criteria for each systematic review are shown in Table 3.1.

The complete manuscripts of selected citations were then reviewed in full to reach the final

decision on inclusion or exclusion. Studies were excluded if numerical data assessing pain were

not presented explicitly (e.g. some papers displayed results graphically such that the mean pain

scores could only be estimated from the graph and given this ambiguity were excluded from

further analysis). An attempt was made to contact authors for missing data but in some instances

this was not possible (11) and in others no reply was received (45-47). When duplicate data were

11

published, only the most up to date, larger series were included. Any disagreements about study

eligibility were resolved by consensus. Inter-rater agreement for study selection was assessed

using the kappa statistic (48)

12

Table 3.1. Search terms and selection criteria for the seven systematic reviews and meta-analyses which investigate pain

reduction in outpatient hysteroscopy Topic Search terms Study selection criteria

Medline,

EMBASE,

CINAHL

Cochrane Library Population Intervention Outcome Study Design

Local anaesthesia ‘hysteroscopy’,

‘vaginoscopy’,

‘local anaesthetic’

and associated

Medical Subject

Headings

‘hysteroscopy’,

‘anaesthetic’

Women

undergoing

diagnostic or

operative

hysteroscopy in the

outpatient setting

i.e. without general

anaesthesia

Use of local

anaesthetic for pain

relief during the

procedure (e.g.

intracervical block,

paracervical block,

local anaesthetic

instilled into the

cavity or applied to

the ectocervix - see

figure 2.1),

compared to no

intervention,

placebo, oral

analgesics or

conscious sedation.

Assessment of pain

(primary outcome)

and vasovagal

episodes

(secondary

outcome)

associated with the

procedure

Randomised

controlled trials

Analgesia ‘hysteroscopy’,

‘vaginoscopy’,

‘analges*and

associated Medical

Subject Headings

‘hysteroscopy’,

‘analgesia’

As above. Use of analgesics

for pain relief

during the

procedure

compared to no

intervention or

placebo.

Assessment of pain

associated with the

procedure (primary

outcome) and

medication side

effects (secondary

outcome).

Randomised

controlled trials

13

Table 3.1 continued Topic Search terms Study selection criteria

Medline,

EMBASE,

CINAHL


Conscious sedation ‘hysteroscopy’,

‘vaginoscopy’,

‘conscious

sedation’,

‘sedative’,

‘sedati*’, ‘pain’

and associated

Medical Subject

Headings

‘hysteroscopy’,

‘sedation’

As above Use of conscious

sedation for pain

relief during the

procedure

compared to no

intervention or

placebo.

Assessment of pain

associated with the

procedure (primary

outcome) and

medication side

effects (secondary

outcome).

Randomised

controlled trials

Cervical

preparation

‘hysteroscopy’,

‘vaginoscopy’,

‘cervical ripening’,

’laminaria’,

’progest*’,

‘prostaglandin’,

‘oestrogen’

‘cervical

preparation’ and

associated Medical

Subject Headings

‘hysteroscopy’,

‘cervical’

As above

Use of cervical

preparation prior to

the procedure,

compared to no

intervention, or

placebo.

Assessment of pain

associated with the

procedure.

Randomised

controlled trials

Vaginoscopic

approach

‘hysteroscopy’,

‘vaginoscopy’,

‘vaginoscop*’, ‘no-

touch’ and

associated Medical

Subject Headings

‘hysteroscopy’,

‘vaginoscopy’,

‘vaginoscopic’,

‘no-touch’

As above

Comparison of the

vaginoscopic

technique versus

hysteroscopy using

a vaginal

speculum.

Assessment of pain

associated with the

procedure and

feasibility

(secondary

outcome).

Randomised

controlled trials

* is used in the search terms to identify all possible suffixes e.g. analges* will identify analgesia, analgesic, and analgesics

14

Table 3.1 continued

Topic Search terms Study selection criteria

Medline,

EMBASE,

CINAHL


Distension media

‘hysteroscopy’,

‘vaginoscopy’,

‘vaginoscop*’,

‘(uter* AND

disten*)’,

‘distension media’,

‘sodium chloride’,

‘normal saline’,

carbon dioxide’,

‘dextran’,

‘mannitol’ and

associated Medical

Subject Headings

‘hysteroscopy’,

‘distension’

As above

Comparison of the

use of carbon

dioxide versus

another distending

medium for the

outpatient

hysteroscopy.

Assessment of pain

associated with the

procedure.

Randomised

controlled trials

Type of

hysteroscope

‘hysteroscopy’,

‘vaginoscopy’,

‘flexible’, ‘rigid’,

‘pain’ and

associated Medical

Subject Headings

‘hysteroscopy’,

‘flexible’, ‘rigid’

As above

Comparison of the

use of flexible

versus rigid

hysteroscope

Assessment of pain

associated with the

procedure.

Randomised

controlled trials

* is used in the search terms to identify all possible suffixes e.g. analges* will identify analgesia, analgesic, and analgesics

15

Data extraction

For each systematic review data were extracted from the selected studies using a piloted data

extraction form (see Appendix 2 for an example). Data were extracted independently by two

reviewers. Data were collected from each trial for study quality (the confidence that the trial

design, conduct, and analysis has minimized or avoided biases in its treatment comparisons) (49),

the intervention, technical aspects of the outpatient hysteroscopy, assessment of pain and for the

relevant secondary outcomes (feasibility, vasovagal episodes, effectiveness, side-effects).

Jadad’s scoring method (Appendix 3) which allowed a quality score on a 5-point scale to be

calculated (50;51) was used to assess the quality of the selected studies in the systematic reviews

of local anaesthetic, analgesia, conscious sedation and cervical preparation. Papers that scored

more than three points were considered to be of high quality. For the systematic reviews of

vaginoscopy, distension media and type of hysteroscope the scoring system was adapted to allow

for the fact that blinding would not have been possible in the studies.

Further information regarding the seven different topics reviewed is detailed in the sections

below.

Local anaesthetic

Table 3.2 shows the quality assessment of the studies selected for use in the systematic review of

the use of local anaesthetic for outpatient hysteroscopy.

16

Table 3.2. Methodological quality assessment of the studies included in the systematic

review of use of local anaesthetic for outpatient hysteroscopy Study Randomised ±1 Double

blind?

±1 Withdrawals

and dropouts

Total Quality (>3 =

high)

Al-sunaidi(52) 1 1 0 0 1 3

Low

Bellati(53) 1 0 0 0 1 2

Low

Broadbent(54) 1 1 1 1 1 5

High

Cicinelli 1997 (9) 1 1 1 1 1 5

High

Cicinelli 1998 (10) 1 1 1 1 1 5

High

Costello(55) 1 1 1 1 1 5

High

Davies (56) 1 1 1 1 1 5

High

Esteve (57) 1 0 1 1 1 4 High

Finikiotis (58) 1 -1 0 0 1 1 Low

Giorda (59) 1 1 0 0 1 3

Low

Guida (60) 1 1 0 0 1 3

Low

Kabli (61) 1 1 0 0 1 3

Low

Lau 1999(12) 1 1 1 1 1 5

High

Lau 2000 (62) 1 1 1 1 1 5

High

Makris (63)

1 0 0 0 1 2

Low

Sagiv (33) 1 1 0 0 1 3 Low

Shankar(64) 1 1 0 0 1 3 Low

Soriano(65) 1 1 1 1 1 5

High

Vercellini (66) 1 1 0 0 1 3

Low

Wong (67) 1 1 1 1 1 5

High

See Appendix 3 for explanation of scoring method.

17

Studies varied in how pain was assessed. Some studies gave an overall pain score for the

procedure. Others scored each of the steps separately (i.e. tenaculum application, administration

of anaesthetic or placebo, insertion of the scope, inspection of uterine cavity, during the biopsy

and at intervals after the end of the procedure). When an overall pain score was given, this was

used for the meta-analysis but when the individual steps were scored, and no overall score was

given, the score relating to inspection of the uterine cavity was deemed most appropriate to use

rather than the scores for other aspects of the procedure (e.g. cervical dilatation, endometrial

biopsy). When scores were only given after the procedure, the most immediate score was used.

Vasovagal reactions during outpatient hysteroscopy are reported to occur in 20% of women in the

general population (8). Parasympathetic nerve stimulation during passage of instruments through

the cervical canal has been suggested as the cause of vasovagal episodes. To examine the

incidence of vasovagal episodes in relation to use of local anaesthetics, data were extracted as 2 x

2 contingency tables (occurrence versus non-occurrence).

18

Analgesia

Table 3.3 shows the quality assessment of studies that were selected for use in the systematic

review of analgesia for pain control during outpatient hysteroscopy.

Table 3.3 Methodological quality assessment of the studies included in the systematic

review analgesia for pain control during outpatient hysteroscopy Study Randomised ±1 Double

blind?

±1 Withdrawals

and dropouts

Total Quality (>3 = high)

Bellati (53) 1 0 0 0 1 2

Low

Caligiani (68) 1 0 0 0 1 2

Low

Floris (45) 1 1 1 1 1 5

High

Lin (69) 1 1 0 0 1 3

Low

Nagele (46) 1 0 1 1 1 4

High

Tam (70) 1 1 1 1 1 5

High


Again, studies varied in how pain was assessed so the same rules were applied to this review as

for the local anaesthetic review (see page 17).

Side effects are an important consideration when administering drugs and it must be established

whether the benefit of the medication outweighs any adverse effects experienced by the patient,

hence the incidence of reported side effects was collected.

19

Conscious sedation

Only one study was selected from the abstracts of studies which looked at the use of conscious

sedation for outpatient hysteroscopy. This was assessed as a low quality study (see Table 3.4).


review of conscious sedation for pain control during outpatient hysteroscopy Study Randomised ±1 Double

blind?

±1 Withdrawals

and dropouts

Total Quality (>3 = high)

Guida (60) 1 1 0 0 1 3 Low


Cervical preparation

The quality assessment of the studies selected for the systematic review of cervical preparation is

detailed in .Table 3.5.


review of the effect of cervical preparation on pain during outpatient hysteroscopy

Study Randomised ±1 Double

blind? ±1

Withdrawals and

dropouts Total

Quality (>3 = high)

Atay (71) 1 0 0 0 1 2 Low

Ben-Chetrit (72) 1 1 1 1 1 5 High

Da Costa (73) 1 1 1 1 1 5 High

Singh (74) 1 1 0 0 1 3 Low

Valente (75) 1 1 1 1 1 5 High

Waddell (76) 1 1 1 1 1 5 High


20

Data regarding pain were reported in an inconsistent manner so were collected for the most

appropriate time point reported. When an overall pain score was given, this was used but when

the individual steps were scored, and no overall score was given, the score relating to inspection

of the uterine cavity was deemed most appropriate to use rather than the scores for other aspects

of the procedure (e.g. cervical dilatation, endometrial biopsy). When scores were only given

after the procedure, the most immediate score was used. The evidence is unclear as to whether

the dilatory effect of prostaglandins is apparent in a non-pregnant cervix, moreover, any potential

benefit has to be weighed against unpleasant side effects (nausea, vomiting, excessive bleeding)

and the costs associated with prostaglandin use. Therefore data regarding the effect on dilatation

and the presence of side effects were collected in 2 x 2 contingency tables.

Vaginoscopy

The quality assessment of the data for the systematic review of the vaginoscopic technique is

shown in Table 3.6.

21


review of the effect on pain of the vaginoscopic approach to outpatient hysteroscopy Study

Randomisation

sequencea

Allocation

Concealmentb

Follow-upc Total Adequate

Almeida (77)

Adequate Inadequate 100% 2

Garbin (31)


Guida (78)

Adequate Adequate 100% 3

Paschopoulos (79)

Not reported Not reported 100% 1

Sagiv (33)

Adequate Not reported 100% 2

Sharma (32)


aRandomisation adequate if computer generated random number sequence. b Concealment considered adequate if third party e.g. nursing staff. Inadequate if sealed envelopes used. c If the total number of patients entering the trial are accounted for in follow-up it was considered 100%. This is because if the

patients have not had the procedure they would be unable to contribute to the results but the authors are able to explain why data

are missing for these patients.

The rules regarding pain score data were applied as in the previous reviews. Data regarding

feasibility of the procedure were extracted as 2 x 2 contingency tables (successful versus failed

procedures).

Distension media

The quality assessment of the data for the systematic review of distension media is shown in

Table 3.7

22


review of the effect of the distension medium used on pain during outpatient hysteroscopy. Study

Randomisation

sequencea

Allocation

concealmentb

Follow-upc Total Adequate

Brusco (80)

Inadequate

Inadequate

100%

1

Lavitola (81)

Adequate Adequate 100% 3

Litta (82)

Inadequate Inadequate 100% 1

Nagele (83)


Paschopoulos (35)


Shankar (64)





The majority of studies gave an overall pain score for the procedure. One scored the steps

separately (i.e. insertion of the endoscope, inspection of the uterine cavity, during endometrial

biopsy and at intervals after the end of the procedure) so the score relating to inspection of the

uterine cavity was used as it seemed the most appropriate.

Data were collected regarding image quality as this may be adversely affected by the type of

distension medium used. Normal saline has a higher refractive index than air which causes

magnification and reduces the visual field (35). Carbon dioxide does not create a lavage and so

blood, mucus and bubbles may obscure the image. The different mechanisms of administration

(insufflators for carbon dioxide and pressure bags for normal saline) may affect the length of the

procedure which prompted the collection of data regarding procedural time. Data regarding

shoulder tip pain (a common side effect of gas insufflations caused by gas leaking from the

23

fallopian tubes and irritating the phrenic nerve) and vasovagal episodes (a common side effect of

hysteroscopy) were also extracted.

Type of hysteroscope

The quality assessment of the data for the systematic review of flexible versus rigid

hysteroscopes is shown in Table 3.8.


review of the effect on pain of the type of hysteroscope used for outpatient hysteroscopy Study

Randomisation

sequencea

Allocation

Concealmentb

Blindingc

Follow-upd Total

Adequate

Baxter (84)

Adequate

Not reported

Single blind

100%

3

Unfried (85)

Adequate Not reported Not reported 100% 2




Pain data were collected and data regarding failed procedures were collected in 2x2 contingency

tables.

24

Data Synthesis

Meta-analysis of pain data was conducted in the systematic reviews of local anaesthetic,

distension media and use of the vaginoscopic approach. The standardised mean difference (SMD)

was used because it allowed comparison of outcome data from studies using different scales to

quantify pain (43). Heterogeneity was assessed by examining forest plots and the I2

statistic,

which if greater than 75% suggests considerable heterogeneity (43). Studies were weighted by

the inverse of the variance and random effects models were used as standard as they give

conservative estimates of effect (43). This method was also used to assess the secondary outcome

of procedure time in the systematic review of distension media. Data assessing pain with the use

of analgesics, conscious sedation, cervical preparation and type of hysteroscope were not suitable

for meta-analysis.

Meta-analysis was possible for secondary outcomes in the systematic reviews of local

anaesthetic, cervical preparation, vaginoscopy and distension media. For dichotomous outcomes

the Peto method was used due to a low incidence of outcome events in the studies (86). Analyses

were performed using RevMan software (87).

In the systematic review of local anaesthesia, subgroup meta-analysis was performed for data

grouped according to method of local anaesthetic administration (intracervical, paracervical,

transcervical and topical) because it was felt that the different methods were not directly

25

comparable. Meta-regression analysis (88) was then used to explore if one of the four types of

local anaesthetic techniques was superior. Meta-regression was performed using Stata (89).

Subgroup analysis was not performed in the six other systematic reviews.

26

CHAPTER 4

RESULTS OF THE SEVEN SYSTEMATIC REVIEWS AND

META-ANALYSES WHICH INVESTIGATE PAIN REDUCTION

IN OUTPATIENT HYSTEROSCOPY

Results of the systematic review and meta-analysis of local anaesthesia for

pain control during outpatient hysteroscopy

Study Selection, Details and Quality

The literature search yielded 245 citations. Reviewing the reference lists yielded two further

citations. Of these, 20 studies were considered eligible for inclusion in the review (Figure 4.1).

The inter-rater reliability for the study selection was very good (kappa=0.9). Details of the study

populations, intervention, outcome assessment and data reporting are shown in Tables 4.1a- 4.1c.

The quality of the studies varied with deficiencies in randomisation and blinding (Figure 4.2).

27

Figure 4.1 Study selection process for systematic review of local anaesthetic for pain relief

during outpatient hysteroscopy

28

Of the 20 selected studies, 18 reported data comparing local anaesthetic to placebo or nothing.

One of these studies also reported data for a third randomised group of patients who received

opiate analgesia (tramadol) (53). Of the remaining two studies, one compared use of local

anaesthetic to conscious sedation (midazolam) (60) and the other compared different local

anaesthetic regimens (paracervical injection versus uterosacral ligament injection) (58).

Of the 18 papers reporting data for pain relief, three were excluded from meta-analysis; two

because data were reported as the median value (56) or the mean but without standard deviation

or standard error (63) precluding calculation of the SMD and another because of differences in

intervention between the groups in addition to the use of local anaesthetic (33). The majority of

the papers used continuous visual analogue scales (VAS) to assess pain, other studies used

ordinal numerical or descriptive scales (Tables 4.1a-4.1c).

29

Tables 4.1a-c Characteristics of the studies included in the systematic review of use of local anaesthetic during outpatient

hysteroscopy, subgrouped according to distension media.

Table 4.1a Studies using carbon dioxide as the distension medium Study

Participants

Intervention

Comparison Outcome measure Data reported

Bellati (53)

(Study written

in Italian,

abstract in

English also)

Women undergoing diagnostic

out-patient hysteroscopy and

endometrial biopsy.

Intracervical injection of

4ml 2% mepivicaine, 5

minutes before the

procedure. N=40

2 groups:

1. Tramadol

100mg i.m. 50

mins pre-

procedure.

N=40

2. Nil. N=40

Ordinal score 0-20 during

the hysteroscopy.

Mean and

Standard

deviation

calculated from

raw data.

Broadbent (54)


out-patient hysteroscopy for

abnormal uterine bleeding.

Exclusions: patients who were

unable to tolerate the procedure.


10ml 1% lignocaine with

1:200,000 adrenaline, at

least 5 minutes before

the procedure. N=49

Intracervical injection

with 10ml 0.9% saline, at


the procedure. N=48

Pain defined by selecting a

category from none, mild,

moderate and severe. Graded

before, during, immediately

and 30min after the

procedure.

Mean and

Standard

deviation

calculated by

assigning a

numerical value

to the groups.

Cicinelli 1997

(9)

Post-menopausal women

undergoing diagnostic

hysteroscopy and endometrial

biopsy because of endometrial

bleeding.

2ml 2% mepivicaine

injected transcervically

through the os into the

uterine cavity 5 minutes

before the procedure.

N=40

2ml 0.9% saline injected

transcervically through

the os into the uterine

cavity 5 minutes before

the procedure. N=40

VAS 0-20 completed before,

during and 15 minutes after

the procedure and during the

endometrial biopsy.

Mean and

standard

deviation

reported.

Cicinelli 1998

(10)

Post-menopausal women


hysteroscopy and endometrial

biopsy because of endometrial

bleeding.

Paracervical block of

10ml 1.5% mepivicaine

10minutes before the

procedure. N=36

Paracervical injection of

10ml 0.9% saline

10minutes before the

procedure. N=36

VAS 0-20 completed before,

during and 15 minutes after

the procedure and during the

endometrial biopsy.

Mean and

standard

deviation

reported.

30

Table 4.1a continued

Study

Participants

Intervention


Costello (55)

Women referred for out-patient

hysteroscopy

Scope passed into

cervical os until ‘snug’.

5ml 2% lignocaine was

then injected through the

operating channel of the

scope. 2 minutes then

passed before the

procedure continued.

N=49

Scope passed into

cervical os until ‘snug’.

5ml 0.9% saline was then

injected through the

operating channel of the

scope. 2 minutes then

passed before the

procedure continued.

N=50

VAS 0-10cm to score pain

during the procedure.

Mean and

standard

deviation

reported.

Davies (56)

Women requiring out-patient

hysteroscopy.

Exclusions: known sensitivity to

lignocaine, epilepsy, significantly

impaired respiratory or cardiac

function, liver disease, treatment

with tricyclic antidepressants or

monoamine oxidase inhibitors.

10% lignocaine sprayed

onto the endocervix and

through the cervical os

into the uterine cavity, 10

sprays in total. N=60

Placebo sprayed onto the

endocervix and through

the cervical os into the

uterine cavity, 10 sprays

in total. N=60

VAS 10cm to score pain as

the tenaculum was applied,

the nozzle of the spray

inserted into the canal,

insertion of the scope,

during the procedure, during

the biopsy and 5 minutes

after the end of the

procedure.

Median VAS and

interquartile

ranges reported.

Esteve (57)

Women attending for out-patient

hysteroscopy.


8ml 2% lignocaine.

N=34


8ml 0.9% saline. N=28

VAS 0-10 cm to score pain

during the hysteroscopy,

during the biopsy, at the end

of the procedure and 30

minutes after the end of the

procedure.

Mean and

standard

deviation

reported.

31


Study

Participants

Intervention


Giorda (59)

All post-menopausal women

referred for diagnostic out-patient

hysteroscopy.

Exclusions: patient refused to

partake, allergy to anaesthesia,

previous hysteroscopy, and

previous severe vagal reaction to

a blind endometrial biopsy.


20ml 1% mepivicaine at


the procedure.

Hysteroscopy performed

with a 5mm diameter

scope. N=121

2 groups:

1. No paracervical

injection.

Hysteroscopy

performed with

a 5mm scope.

2. No paracervical

injection.

Hysteroscopy

performed with

a 3.5mm

diameter scope.

N=119

Visual numerical rating

scale ranging from 0 to 10 to

score pain during the

procedure only (patients

who received a paracervical

block were asked to discount

the pain from the injection

from their assessment.)

Mean reported.

Standard

deviation

calculated from

standard error.

Lau 1999 (12)


outpatient hysteroscopy for

abnormal uterine bleeding.


10ml 2% lignocaine 5

minutes before the

procedure.

N=49


10ml 0.9% saline 5

minutes before the

procedure.

N=50

VAS 10cm used to score the

pain when the tenaculum

was applied, after the

paracervical injection, at

hysteroscopy insertion,

during hysteroscopy, after

endometrial biopsy and 30

minutes after the procedure.

Mean and

standard

deviation

reported.

Lau 2000 (62)

Women scheduled for diagnostic


5ml 2% lignocaine

instilled transcervically

into the uterine cavity.

N=45

5ml 0.9% saline instilled

transcervically into the

uterine cavity. N=44

VAS 10cm used to score the

pain when the tenaculum

was applied, after the

paracervical injection, at

hysteroscopy insertion,

during hysteroscopy, after


minutes after the procedure.

Mean and

standard

deviation

reported.

32


Study

Participants

Intervention


Makris (63)


outpatient hysteroscopy ±

endometrial biopsy.


1-3ml 3% mepivicaine, 3

minutes before the

procedure. N=100


1-3ml 0.9% saline, 3

minutes before the

procedure. N=100

Ordinal scale 0-10. Patients

asked to rate pain

experienced during

hysteroscopy and at 30 and

60 minutes after the

procedure by circling one of

the numbers.

Mean reported.

Unable to

calculate

standard

deviation.

Wong (67)

Women referred for investigation

of abnormal uterine bleeding or

suspected endometrial pathology.

Exclusions: women who spoke a

dialect (study carried out in

China) or had other

communication problems.

4ml of 2% lignocaine

rubbed over the cervix

for 20 seconds

immediately before the

hysteroscopy. N=250

4ml of KY Jelly

(Johnson and Johnson

Medical, UK) rubbed

over the cervix for 20

seconds immediately

before the hysteroscopy.

N=250

Patients asked to grade the

severity of pain at 1 minute

intervals using the PPI scale.

The mean pain score, peak

pain score and overall pain

score were all calculated as

were mean pain scores for

each of the individual steps

of the procedure.

Mean and

standard

deviation

reported.

33

Table 4.1b. Studies using normal saline as the distension medium Study

Participants

Intervention


Al-sunaidi (52)

Women undergoing diagnostic out-

patient hysteroscopy for evaluation of

uterine cavity.

Exclusions: women needing operative

hysteroscopy under GA, positive

Chlamydia culture, pregnancy or

allergy to local anaesthetic.


of 2ml 0.5%

bupivacaine and

paracervical injection

of 8ml 0.5%

bupivacaine, 5 minutes


N=42

Intracervical

injection of 2ml

0.5% bupivacaine, 5

minutes before the

procedure. N=42

VAS 0-10, completed

during the procedure

and at 10, 30 and 60

minutes post

procedure.

Mean and standard

deviation reported.

Guida (60)

Women undergoing operative out-

patient hysteroscopy for surgically

treatable lesions associated with

infertility or abnormal uterine

bleeding.

Paracervical injection

of 10ml 1%

mepivicaine. N=82

Conscious sedation

with o.5mg atropine

i.v., 0.25mg fentanyl

i.v. and 2mg

midazolam i.v. N=84

5cm VAS used during,

immediately after, 15

and 60 minutes after

and 24 and 72 hours

after the procedure.

Mean and standard

deviation reported.

Kabli (61)

Infertile women undergoing


Exclusions: women needing operative

hysteroscopy under GA, positive

Chlamydia culture, pregnancy or

allergy to local anaesthetic.


of 2ml 1% lignocaine

and distension media

with 18ml lignocaine

per 250ml saline. N=42

Intracervical

injection of 2ml 1%

lignocaine. N=36

VAS 0-10 used to

score pain after the

hysteroscopy, after

endometrial biopsy and

at 10, 30 and 60

minutes after the

procedure.

Mean and standard

deviation reported.

Sagiv (33)




of 10ml 3%

mepivicaine. N=47

Vaginoscopy

(procedure

performed without a

speculum or

anaesthesia). N=83

VAS 0-10cm used to

score the pain

immediately and 15

minutes after the

hysteroscopy.

Mean and standard

deviation reported.

Shankar (64)

Women with abnormal uterine

bleeding referred by their general

practitioner for diagnostic outpatient

hysteroscopy.

Exclusions: unable to visualize the

cervix or severe cervical stenosis.

Distension media

containing 40ml 2%

lignocaine per 500ml

0.9% saline. N=100

Distension media of

0.9% saline only.

N=100

Pain scored with VAS

0-10, and PPI.

Mean and standard

deviation reported.

34

Table 4.1b continued

Study

Participants

Intervention


Soriano (65)


hysteroscopy for abnormal uterine

bleeding or infertility.

Exclusions: menorrhagia at the time

of the procedure, sensitivity to

lignocaine, epilepsy, significantly

impaired respiratory or cardiac

function and active liver disease.

5% lignocaine sprayed

onto the endocervix

and into the cervical

canal, (3 sprays in

total) 5 minutes before

the procedure. N=62

Placebo sprayed on

the endocervix and

into the cervical

canal, (3 sprays in

total) 5 minutes


N=56

VAS 0-10cm to score

pain experienced


Mean and standard

deviation reported.

35

Table 4.1c Studies using other distension media Study

Participants

Intervention


Studies using 1.5% glycine as the distension medium

Vercellini (66)

Premenopausal (FSH <30mIU/ml)

non-pregnant (negative β-hCG test)

women referred for investigation of

excessive uterine bleeding of ≥ 3

months duration.

Exclusion: genital infection, previous

cervical surgery or hysteroscopy,

severe cardiac disease and known

sensitivity to local anaesthetics.


of 1% mepivicaine

more than 5 minutes


N=87

No anaesthesia.

N=90

10 point visual analogue

scale used to score pain

during the hysteroscopy and

the endometrial biopsy.

Mean and standard

deviation reported.

Distension medium not stated

Finikiotis (58)

Patients referred from general

practitioners and from other

gynaecologists for the investigation

of a variety of gynaecological

complaints.


of 16-20ml 1%

lignocaine. N=60

Uterosacral

injection of 2ml

2% lignocaine

with 1:80,000

adrenaline. N=60

VAS 0-10cm to score pain


Reported as the number of

patient selecting VAS

between 0 and 3.3, 3.4 and

6.3 and 6.4 to 10.0.

Mean and standard

deviation

calculated using

the mean value of

each category.

NB. For consistency the group receiving local anaesthetic (or a combination of anaesthetics) are considered as the intervention group even if that was not the case in the original

study.

GA= general anaesthetic, VAS= visual analogue scale, i.m.= intra muscular, i.v.= intra vascular, HR= heart rate, BP= blood pressure, PPI= Present pain intensity scale (verbal

descriptors of pain ranked from 0-5 on a numeric scale).

36

Figure 4.2 Jadad Quality Assessment of Studies Using Local Anaesthetic for Out-patient

Hysteroscopy

Results were reported as mean or median pain scores, but for the one study using a descriptive

scale (54), numerical values were applied to each category (none=1, mild=2, moderate=3 and

severe=4) and used to calculate the mean scores and standard deviations (90). One study reported

raw patient data (53) from which the mean and standard deviation were calculated (90). The

populations in the two studies (53;54) for which the mean and standard deviation were calculated

were sufficiently large for them to be approximated to a normal distribution according to central

limit theorem (91). Another study reported the standard error (59) which was converted into the

standard deviation (92).

37

Nine of the selected studies had data on vasovagal episodes. Four of the studies reported

vasovagal attacks according to a strict definition based upon heart rate, blood pressure and

symptoms (9;10;12;62), four of them reported vasovagal symptoms (e.g. faintness, nausea,

pallor) (59;60;63;64) and one reported a vasovagal attack in the complications but did not give

any a-priori definition of symptoms or signs (65).

Effect of local anaesthetic on pain

Meta-analysis of 15 studies showed that the use of local anaesthetic reduced the amount of pain

experienced during outpatient hysteroscopy (SMD = -0.54, 95%CI -0.86 to -0.23, I2

= 91%)

(Figure 4.3). Meta-analysis of the studies sub-grouped according to quality found that both the

poor and the high quality studies demonstrated a significant benefit of using local anaesthetic for

outpatient hysteroscopy (SMD = -0.77, 95%CI -1.45 to -0.08, I2

= 95% and SMD = -0.43, 95%CI

-0.73 to- 0.12, I2

= 83% respectively) (Figure 4.3).

38

Figure 4.3 Forest plot showing the results of meta-analysis of studies that examine the use of

local anaesthetic for reducing pain during out-patient hysteroscopy. Results overall and

sub-grouped according to method of administration and quality.

When divided into subgroups there were three studies examining intracervical injection

(53;54;57), five which used paracervical injection, (10;12;52;59;66), five that used transcervical

application (topical into the uterine cavity) (9;55;61;62;64) and two that applied the anaesthetic

topically (topical to the cervix only) (65;67).

39

The use of an intracervical injection of local anaesthetic significantly reduced pain during out-

patient hysteroscopy (SMD = -0.36, 95%CI -0.61 to -0.10, I2

= 0%) (Figure 4.3). This finding

however, contrasted with a single study included in the review but not in the meta-analysis

because of insufficient data (63), which found no significant effect of intracervical local

anaesthetic on pain relief for out-patient hysteroscopy. To examine this conflicting result, a

sensitivity analysis was performed excluding from the meta-analysis the study where categorical

data had been transformed (54). No significant reduction in pain was observed with intracervical

injection (SMD =-0.35, 95% CI -0.82 to 0.12, I2

=48%) (Figure 4.4).

Figure 4.4 Forest plot showing the results of meta-analysis of studies that examine the use of

intracervical injection without the study whose data were transformed

The use of paracervical injection was associated with a significant reduction of pain during out-

patient hysteroscopy (SMD = -1.28, 95%CI -2.22 to -0.38, I2

=97%) (Figure 4.3).

The use of topically administered local anaesthetic did not ameliorate pain during outpatient

hysteroscopy. Specifically, transcervical local anaesthetic was not found to significantly reduce

the amount of pain experienced during hysteroscopy (SMD= -0.11, 95%CI -0.31 to 0.10, I2

=27%) (Figure 4.3). Similarly, there was no significant alleviation of pain when local anaesthetic

40

was applied topically to the cervix (SMD = -0.32, 95%CI -0.97 to 0.33, I2

= 90%) (Figure 4.3),

although meta-analysis demonstrated substantial heterogeneity. A further study included in the

review that could not be used for the meta-analysis because it reported median VAS scores,

showed no significant difference between topical cervical local anaesthetic and placebo for the

hysteroscopy, but it did show a significant reduction in pain in the local anaesthetic group during

application of a cervical tenaculum (p=0.005) (56).

A further meta-analysis was performed to compare injectable administration of local anaesthetic

(intracervical and paracervical) against topical application (transcervical to uterine cavity and

topical to the cervix). This showed a benefit of using injectable local anaesthetics (SMD = -0.92,

95% CI -1.51 to -0.33, I2

= 94%) but not topical ones (SMD =-0.17, 95%CI -0.38 to 0.03, I2

=

62%). (Figure 4.5). Meta-regression analysis showed that paracervical injection was significantly

more effective than the other anaesthetic modalities in reducing the pain of diagnostic outpatient

hysteroscopy (p = 0.048).

41

Figure 4.5 Forest plot showing the results of meta-analysis of studies that examine the use

anaesthetics sub-grouped into injectable and topical application

A single study compared two methods of cervical block (58) and found no significant difference

in pain between a paracervical and an uterosacral ligament local block p<0.65. Two studies

compared local anaesthetic to other medication (53;60). The first compared intracervical local

anaesthetic to a control group (data used in meta-analysis) and to intramuscular injection of

100mg tramadol. Tramadol was significantly better at reducing the amount of pain experienced

during hysteroscopy (p=0.001) compared to intracervical block (53). The second study compared

paracervical injection of local anaesthetic to the use of conscious sedation for operative

hysteroscopy and found no significant difference in the pain experienced between the two groups

(60).

42

Effect on vasovagal episodes

There was no significant difference in the incidence of vasovagal episodes between local

anaesthetic and control (nil, normal saline, placebo, conscious sedation) groups p=0.09 (Figure

4.6), regardless of how a vasovagal reaction was defined.

Figure 4.6 Forest plot showing the results of meta-analysis examining the incidence of

vasovagal episodes in studies examining the use of local anaesthetic for out-patient

hysteroscopy

43

Results of the systematic review of analgesia for pain control during outpatient

hysteroscopy


The analgesia literature search yielded 185 citations. Of these, 6 studies were considered eligible

for inclusion in the review (Figure 4.7). The inter-rater reliability for the study selection was

good (kappa 0.67).

Figure 4.7 Study selection process for the systematic reviews of the effect of analgesia on

pain during outpatient hysteroscopy.

Details of the study populations, intervention, outcome assessment and data reporting are shown

in Table 4.2.

44

Table 4.2. Characteristics of the selected studies included in the systematic reviews of analgesia for pain control during

outpatient hysteroscopy Study

Participants

Intervention

Comparison Outcome measure Data

reported

Bellati (53)

(Study

written in

Italian,

abstract in

English also)

Women undergoing

diagnostic out-patient

hysteroscopy and

endometrial biopsy.

2 groups:

1. Tramadol 100mg i.m. 50

mins pre-procedure. N=

40

2. Nil. N=40.

Rigid 5mm hysteroscope and carbon

dioxide for distension.

Intracervical injection of 4ml

2% mepivicaine, 5 minutes

before hysteroscopy with a

rigid 5mm hysteroscope and

carbon dioxide for

distension. N=40

Rigid 5mm hysteroscope and

carbon dioxide for

distension.

Ordinal score 0-20

during the hysteroscopy.

Mean and

Standard

deviation

calculated

from raw

data.

Caligiani (68)

(Study

written in

Italian,

abstract in

English also)

Women undergoing

diagnostic out-patient

hysteroscopy and

endometrial biopsy to

investigate post-

menopausal bleeding.

Intracervical injection of 4ml 3%

mepivicaine, 5 minutes before the

procedure and ketorolac 30mg i.m. and

atropine 0.5mg i.m. 45 minutes before

the procedure. Hysteroscopy not

described. N=12.

Intracervical injection of 4ml

3% mepivicaine, 5 minutes

before the procedure and

atropine 0.5mg i.m. 45

minutes before the

procedure. Hysteroscopy

not described. N=12.

VAS 0-10cm to score

pain at insertion of the

speculum, during the

hysteroscopy, during


minutes after the end of

the procedure.

Mean and

standard

error.

Floris (45)

Perimenopausal women

undergoing outpatient

diagnostic hysteroscopy

and endometrial biopsy.

Exclusions: known

uterine malformation,

previous uterine

surgery.

Tramadol 100mg i.v. over 20 minutes,

followed 30 minutes later by the

outpatient hysteroscopy with a 2.9mm,

rigid, 30 degree fore-oblique

hysteroscope using carbon dioxide as the

distension medium. N = 25.

Placebo infusion i.v. over 20

minutes, followed 30

minutes later by the


with a 2.9mm, rigid, 30

degree fore-oblique

hysteroscope using carbon

dioxide as the distension

medium. N = 25.

VAS 0-10cm to score

pain during and 15


the procedure.

ACTH and cortisol levels

were also assayed from

blood samples pre and

post procedure.

Mean

reported

graphically,

no values

given.

45

Table 4.2 continued

Study

Participants

Intervention

Comparison Outcome measure Data

reported

Lin (69)

Pre- and post-

menopausal women


hysteroscopy to

investigate abnormal

bleeding, infertility and

suspected intracavity

lesion.

Buprenorphine 0.2mg sublingually, 40

minutes before the hysteroscopy, which

was performed a 3.1mm flexible

hysteroscope, using 5% dextrose as the

distension medium. N= 80.

Placebo tablet, sublingually,

40 minutes before the

hysteroscopy, which was

performed with a 3.1mm

flexible hysteroscope, using

5% dextrose as the

distension medium. N=84.

VAS 0-10cm to score the

worst pain experienced


Mean and

standard

deviation.

Nagele (46)

Pre- and post-

menopausal women


hysteroscopy.

Exclusions: sensitivity

to NSAID’s and

prostaglandin synthesis

inhibitors, medication

that interacts with

NSAID’s, inflammatory

bowel disease,

porphyria and those

unable to consent.

Mefenamic acid 500mg orally, 1 hour

pre-procedure. Hysteroscopy with 4mm,


hysteroscope with a 5mm sheath using

carbon dioxide as the distension medium.

N= 49.

Placebo orally, 1 hour pre-

procedure. Hysteroscopy

with 4mm, rigid, 30 degree

fore-oblique hysteroscope

with a 5mm sheath using

carbon dioxide as the

distension medium. N- 46.

Women asked to score

the worst pain

experienced during the

hysteroscopy and at 30

and 60 minutes after the

procedure using a 4 point

ordinal scale (0=none,

1=mild, 2=moderate,

3=severe).

No values

given for pain

during the

procedure,

results

expressed

graphically.

Pain after the

procedure

reported as

medians with

interquartile

ranges

Tam (70)

Pre- and post-

menopausal women


diagnostic

hysteroscopy.

Diclofenac 50mg, orally, 1-2 hours pre-

procedure. Hysteroscopy with 2.7mm,


hysteroscope with a 5mm sheath and

carbon dioxide for distension. N=92.

Placebo tablet, orally, 1-2

hours pre-procedure.

Hysteroscopy with 2.7mm,


hysteroscope with a 5mm

sheath and carbon dioxide

for distension. N=89.

VAS (length not stated)

to score pain before the

procedure, when

tenaculum applied,

insertion of

hysteroscope,

examination of the

uterine cavity, after

biopsy and 30 minutes

post procedure.

Mean and

standard

deviation.

46

The quality of the studies varied with deficiencies in randomisation and blinding (Figure 4.8).

Figure 4.8. Methodological quality assessment of studies using analgesia for outpatient

hysteroscopy

Of the 6 selected analgesia studies, three examined the use of opiate drugs (45;53;69) and three

examined non-steroidal anti-inflammatory drugs (NSAIDs) (46;68;70). All six studies examined

diagnostic outpatient hysteroscopy. Four studies reported data comparing analgesia to placebo

(45;46;69;70), one compared ketorolac and intracervical block to intracervical block alone (68)

and the final study compared tramadol to intracervical block (53). Two of the selected studies

reported their data graphically without any specific numerical values (45;46); the authors were

contacted to try and obtain the values but no replies were received. Four of the papers used

47

continuous visual analogue scales (VAS) to assess pain, and two used ordinal numerical or

descriptive scales (Table 4.2).

Effect of analgesics on pain

Meta-analysis was not performed as the studies used different drugs, in different doses, making

them non-comparable. When divided into medication subgroups, there were three studies

examining the use of opiate analgesics (45;53;69) and three which studied NSAID’s (46;68;70).

Two of the opiate studies examined the use of 100mg tramadol administered approximately 50

minutes before the outpatient hysteroscopy, one study giving it intramuscularly (53) and the

second giving it as an intravenous infusion (45). The first study found that the women who had

received tramadol had significantly less pain at the end of the procedure than women in the

intracervical block group and the women who received no medication ( p =0.001 and p<0.001

respectively) (53). Although this was a low quality study the result was supported by those from

the second, high quality study that reported significantly lower pain scores in the tramadol group

when compared to placebo both during (p<0.012) and 15 minutes after ( p<0.008) the procedure

(45). The third opiate study examined the use of sublingual buprenorphine 0.2mg, 40 minutes

before the procedure, versus placebo. There was no significant pain reduction with the use of

buprenorphine overall and when stratified for menopausal status and parity (69).

The three NSAID studies all examined different drugs. The first, high quality study compared the

use of 500mg of oral mefenamic acid, one hour prior to the procedure, to placebo. It found that

mefenamic acid did not significantly reduce the pain of the hysteroscopy, however it did

significantly reduce the pain experienced at 30 minutes (p<0.01) and 60 minutes (p<0.05) post-

48

procedure. The second high quality study examined oral diclofenac 50mg, given one to two hours

before the procedure (70). Pain was assessed at a number of points during the procedure (see

Table 4.2) but no significant difference between the intervention and control groups was found (p

values not reported) (70). The final NSAID study was low quality. It assessed the use of oral

ketorolac 30mg with an intracervical block compared to intracervical block alone and found that

the addition of ketorolac significantly reduced pain at all stages of the procedure (no p values

reported) (68).

Side effects

Two of the opiate studies reported side effects (45;70). The tramadol study found no significant

difference between the groups in incidence of nausea, vomiting or bradycardia (45). Conversely,

in the buprenorphine study there was a high incidence of adverse effects (nausea, vomiting,

drowsiness) in the intervention group (38.8%) and none in the control group (69).

The diclofenac study was the only NSAID study to report side effects (70). Two patients (2.2%)

in the diclofenac group reported side effects (one reported epigastric pain and one reported a

rash) which were mild and self-limiting (70).

49

Results of the systematic review of conscious sedation for pain control during



The conscious sedation literature review yielded 39 citations with just one being eligible for

inclusion in the review (Figure 4.9).

Figure 4.9 Study selection process for the systematic reviews of the effect of conscious

sedation on pain during outpatient hysteroscopy

50

There was complete agreement between the reviewers and so the kappa score was 1. According

to the Jadad quality assessment this paper was of low quality, however this was due to the study

not scoring any points for blinding. It would be very difficult to blind conscious sedation versus

intracervical injection and so blinding was removed as a quality consideration. The paper

received full marks for randomisation and follow-up of withdrawals and dropouts, thus was

ultimately deemed to be of high quality.

The conscious sedation study (60) compared midazolam to a paracervical anaesthetic block for

operative outpatient hysteroscopy using a bipolar electrode. A VAS marked with pain

descriptions was used to evaluate the pain experienced and the results were reported as means

and standard deviations. The trial examined the use of conscious sedation immediately before

operative outpatient hysteroscopy (polypectomy, myomectomy, septoplasty and adhesiolysis)

using the Versapoint™ bipolar electrosurgical system (Gynecare, Ethicon Inc. Menlo Park, CA,

USA). It compared 0.25 mg of fentanyl i.v., 0.5 mg of atropine and 2mg of midazolam, to

paracervical anaesthesia with 10mL 1% mepivicaine hydrochloride without sedation. There were

no significant differences between local anaesthesia and conscious sedation in terms of pain

control during the procedure, postoperative pain, side effects or patient satisfaction (60).

51

Results of the systematic review of the effect of cervical preparation on pain



The literature search yielded 585 citations. Of these, six studies were considered eligible for

inclusion in the review (Figure 4.10) (71-76). The inter-rater reliability for the study selection

was very good (kappa=0.86).

Figure 4.10 Study selection process for the systematic review of the effect of cervical

preparation on pain during outpatient hysteroscopy

52

Four of the six papers were deemed to be of high quality (72;73;75;76) (Table 4.3). Details of the

study populations, intervention, outcome assessment and data reporting are shown in Table 4.3.

A single study (72) examined the use of oral mifepristone for cervical dilatation in

premenopausal women given 30 hours pre-procedure. The five remaining studies examined the

use of vaginal misoprostol administered between four and 24 hours before the hysteroscopy. The

dose used was 400µg in all but one study (73) which used 200µg. Two studies used only

premenopausal women (71;75), one, only postmenopausal women (73) and the remaining two a

combination of both (74;76) . All studies reported results for pain and five (71-74;76) reported

results for cervical dilatation. One paper used an ordinal scale from zero to ten to assess pain (71)

whilst the other studies all used continuous visual analogue scales (VAS).

Results were reported in a variety of ways (see Table 4.3) that could not be compared, making

meta-analysis impossible. Cervical dilatation was assessed in different ways (adequate if scope

passes into os, need for dilatation, force needed to dilate) and so the results were once again

unsuitable for meta-analysis. Gastrointestinal side effects were reported by two studies (75;76),

bleeding by four studies (71;74-76), cervical lacerations by three studies (71;73;76) and failed

procedures by three studies (72;73;75). These results were meta-analysed.

53

Table 4.3. Methods, interventions and results for the studies included in the systematic review of the effect of cervical

preparation on pain during outpatient hysteroscopy Antiprogestogen Prostaglandins

Premenopausal Premenopausal Postmenopausal Pre and postmenopausal

Ben Chetrit (72)

Atay (71) Valente (75) Da Costa (73) Singh (74) Waddell (76)

Intervention

Mifepristone

200mg PO 30

hours pre

hysteroscopy

Misoprostol 400µg

PV 4 hours pre

hysteroscopy

Misoprostol

400µg PV 6

hours pre

hysteroscopy

Misoprostol 200µg

PV 8 hours pre

hysteroscopy

Misoprostol 400µg

PV 4-6 hours pre

hysteroscopy

Misoprostol 400µg PV 12-24

hours pre hysteroscopy

Control method Placebo Placebo

Placebo Placebo Nil Placebo

Number in

intervention

group vs.

control group

28 vs. 30 22 vs. 21 20 vs. 24 60 vs. 60 50 vs. 50 (only 8

postmenopausal

women in the

intervention group

and 1 in the

control group).

50 vs. 51

Routine use of

cervical

dilatation

No If 7mm resectoscope

sheath or 6 Hegar not

easily passed into the

cervical canal.

No No No Yes

Hysteroscope

size

2.9mm rigid

hysteroscope

4mm rigid

hysteroscope with

30° fore oblique lens

for diagnostic and

7mm resectoscope

for operative

4mm

hysteroscope

with 30° fore

oblique lens

4mm rigid

hysteroscope with

30° fore oblique

lens and 5mm

sheath

4mm rigid

hysteroscope with

30° fore oblique

lens

4mm rigid hysteroscope with 30°

fore oblique lens and 5.5mm

sheath

54

Table 4.3 continued

Ben Chetrit (72) Atay (71) Valente (75) Da Costa (73) Singh (74) Waddell (76)

Quality High Low

High High Low High

Operative or

Diagnostic

Diagnostic Suggests all

operative but not

clear.

Diagnostic Diagnostic Diagnostic Diagnostic

Method of

assessing

cervical

dilatation

Dilatation

measured using

Hegar dilators

before

administration of

mifepristone and

when patient

returned for

hysteroscopy. The

change in Hegar is

reported as mean

(standard

deviation).

Adequate if 7mm

hysteroscope

sheath or Hegar 6

dilator passes into

cervical canal.

Number assessed

adequate in each

group.

Not reported. Need for

dilatation.

Percentages.

Need for

dilatation. Number

(percentage).

Force needed to dilate at 3, 4, 5

and 6mm measured with dilators

attached to a tonometer. Mean

(standard deviation).

Dilatation in

intervention

group vs.

control group

1.28 (1.4) Hegar

vs. 1.06 (1.4)

Hegar

p=0.50

20 vs. 6 adequate

p<0.001

17.2 vs. 20.3%

p=0.66

15 (30%) vs. 11

(22%) p=0.36

3mm= 1.7 (1.7)vs. 1.8 (2.1)

p=0.82

4mm= 2.6 (3.5) vs. 3.0 (4.5) p=

0.86

5mm= 4.3 (6.0) vs. 4.0 (3.1) p=

0.21

6mm= 5.0 (4.2) vs. 7.5 (5.9) p=

0.02

Scores separated into pre and

postmenopausal are all non-

significant (p>0.05)

55

Table 4.3 continued

Ben Chetrit (72) Atay (71) Valente (75) Da Costa (73) Singh (74) Waddell (76)

Method of pain

assessment

100mm VAS Score from 0-10

(0= no pain, 10=

worst menstrual

pain).

0-10cm VAS

with faces

symbolising the

amount of pain.

VAS (length not

reported but

assumed to be

10cm).

10cm VAS Sliding visual analogue scale of 1

to 100.

Timing of pain

assessment

Pain as

hysteroscope

passes through the

cervical os. Mean

(standard

deviation).

Pain during

cervical dilatation.

Mean (range).

Pain during the

hysteroscopy,

biopsy and after

the hysteroscopy.

Number

(percentage),

selecting between

0-5 and 6-10

Pain as scope

passes through

internal cervical

os, as the cervix is

clamped and the

biopsy is taken.

Median.

Pain at the end of

the procedure.

Median (standard

deviation).

Pain assessed during assessment

of cervical dilatation at baseline

(before administration of

misoprostol) and after cervix

dilated to 6mm before

hysteroscopy.

Mean (standard deviation).

Pain score in

intervention

group vs.

control group

33.4mm (23.5)vs.

37.0mm (30.0)

p=0.60

5.1 (4-10) vs. 9.3

(6-10) p<0.05

Selecting 0-5: 7

(53.8%) vs. 11

(47.8%)

Selecting 6-10: 6

(46.2%) vs. 12

(52.2%)

p=0.72

During procedure

medians:

5 vs. 7, p=0.02

Comparing

number in group

selecting Vas>6 :

Clamping p =

0.507

Hysteroscopy

p=0.0132

Endometrial

biopsy p= 0.5919

Presence of pain:

Clamping p = 0.74

Hysteroscopy

p=0.32

Endometrial

biopsy p= 0.19

4.5cm (2.0) vs.

5.0cm (1.8) p=0.03

42.1 (23.1) vs. 57.2 (24.9), p=

0.004

If adjusted for baseline pain: 43.2

(3.7) vs. 55.5 (3.4), p= 0.01

Premenopausal:

Overall p = 0.56 or if adjusted for

baseline pain p= 0.77

Postmenopausal:

Overall p=0.004 or if adjusted for

baseline pain p= 0.006

56

Effect of cervical preparation on pain

Premenopausal women

Mifepristone given 30 hours before hysteroscopy did not reduce the pain of the outpatient

procedure for pre-menopausal women (p=0.60) (72) (Table 4.3).

Two studies examined the use of misoprostol 400µg given vaginally before the hysteroscopy

to premenopausal women. The drugs were administered 4 hours before in one of the studies

(71) and 6 hours before in the other (75). The low quality study (71) found that pain during

cervical dilatation was significantly reduced after the use of prostaglandin compared to

placebo (p<0.05) however the other study which, was considered to be high quality (75),

found no significant reduction in pain during the hysteroscopy with the use of misoprostol (p=

0.72) (Table 4.3).

Postmenopausal women

One study (73) examined the use of misoprostol 200µg given vaginally 8 hours before the

hysteroscopy to postmenopausal women. The median pain scores as the hysteroscope passed

through the cervical os were five in the intervention group and seven in the placebo group

(p=0.02). When the pain severity was assessed by comparing the number of patients scoring

more than six on the VAS (i.e. considerable pain) there were significantly fewer in the

intervention group (p=0.0132). However no significant difference between the groups was

identified when assessing the presence of pain during clamping of the cervix (p= 0.74), during

the examination (p= 0.32) or during the endometrial biopsy (p=0.19) (Table 4.3).

Pre- and postmenopausal women

Two studies included pre- and post-menopausal women in their study populations (74;76).

One of these studies gave misoprostol 400µg vaginally four to six hours before the

57

hysteroscopy (74) and found that pain at the end of the procedure was significantly less in the

intervention group when compared to no medication (p=0.03). This was judged to be a low

quality study due to the lack of blinding. The second study gave the same dose of misoprostol

twelve to twenty-four hours before the procedure and assessed pain after the cervix was

dilated to 6mm (76). Pain was found to be significantly less in the misoprostol group

(p=0.004 and when adjusted for baseline pain score p=0.01). This study subgrouped the

patients according to menopausal status and found that there was a significant reduction in

pain for postmenopausal women given misoprostol (p= 0.004 and when adjusted for baseline

scores p=0.006) but not for premenopausal women (p=0.56, and when adjusted for baseline

scores p=0.77). This was a high quality study (Table 4.3).

Effect of cervical preparation on dilatation of the cervix

Five studies evaluated the effect of cervical priming with antiprogestogens or prostaglandins

on cervical dilatation. No significant differences were seen between priming and placebo as

regards the need for or degree of cervical dilatation. The only significant results were for the

amount of force required to dilate the cervix above 6mm in a mixed pre and postmenopausal

population (76) and the number of premenopausal women requiring dilatation above 6mm

(p=0.001) (71) (Table 4.3).

Side effects

No significant differences in the prevalence of gastrointestinal side effects, unscheduled

vaginal bleeding, cervical laceration or failed procedures were found (Figure 4.11).

58

Figure 4.11 Forest plots showing the results of meta-analysis of adverse effects and failed

procedures when prostaglandins are given prior to outpatient hysteroscopy

59

Results of the systematic review of the effect on pain of the vaginoscopic

approach to outpatient hysteroscopy


The literature search yielded 1167 citations. Of these, 6 studies (31-33;77-79) were

considered eligible for inclusion in the review (Figure 4.12). The inter-rater reliability for the

study selection was very good (=0.81).

Figure 4.12 Study selection process for systematic review of the effect on pain of the

vaginoscopic approach to outpatient hysteroscopy.

Potentially relevant citations identified and screened for

retrieval

Medline, EMBASE and CINAHL 1157 Cochrane Library 10

Total n = 1167

Publications selected for appraisal

n = 6

Citations excluded Inappropriate population, intervention or outcome measure n=1060 Duplicates n= 88

Total n = 1048

Publications excluded Not randomized controlled trial n =11 Duplicates n = 2

Total n =13

Citations retrieved for more

detailed evaluation

Total n = 19

60

Details of the study populations, intervention, outcome assessment and data reporting are

shown in Table 4.4. The quality of the studies was high but with deficiencies in allocation

concealment (Figure 4.13).

Figure 4.13 Methodological Quality Assessment of Studies Examining the effect of the

vaginoscopic approach on pain during outpatient hysteroscopy

All six papers (31-33;77-79) used a visual analogue scale (VAS) to assess the patients’ level

of pain. Of the six papers reporting data for pain experience three reported means and

standard deviations (31;33;77), two reported medians (32;78) and one reported the pain in

categories (79). The data reported as median pain scores could not be used to calculate the

SMD, so were excluded from the analysis. For data reported categorically, the midpoint of

each category and the number of people selecting that category were used to calculate the

mean and standard deviation.

0% 20% 40% 60% 80% 100%

Follow-up >90%

Randomisation sequence

Allocation concealment

Adequate

Inadequate

Unreported

61

Table 4.4. Characteristics of the selected studies included in the systematic review of the effect of the vaginoscopic approach on pain


Study

Participants

Intervention


Almeida (77) Pre- and post-menopausal

women undergoing diagnostic


Exclusions: pelvic

inflammatory disease,

pregnancy, uterine perforation

within the last 30 days, active

uterine bleeding and use of

hormonal vaginal cream.

Patients positioned in lithotomy,

vagina cleaned and the rigid,

2.7mm (3.7mm with the sheath)

hysteroscope introduced into the

vagina. Holding the labia minora

together the vagina was distended

with normal saline and the

hysteroscope steered into the

external cervical os and along the

canal into the uterine cavity. The

hysteroscopy then continued as

normal. Endometrial biopsies were

also taken without the use of a

speculum.

N =91.


Collins speculum inserted,

vagina and cervix cleaned and

the cervix grasped with a

tenaculum. The rigid, 2.7mm

(3.7mm with sheath)

hysteroscope was then inserted

into the external cervical canal

and advanced into the uterine

cavity which was then distended

with carbon dioxide. The


normal and endometrial biopsies

were taken. N=93.

VAS 0-10 with five

faces drawn above the

line indicating the

amount of pain. Scored

during the procedure

and 5, 10, 15 and 20

minutes afterwards.

Mean and

standard

deviation

reported.

Garbin (31) Pre- and post-menopausal



Exclusions: age < 18 years,

current genital infection,

refusal to participate.

Vagina cleaned and the rigid,

2.7mm (3.5mm with the sheath)

hysteroscope introduced into the

vagina, which was distended with

normal saline so that the

hysteroscope could be steered into

the external cervical os and along

the canal into the uterine cavity.

The hysteroscopy then continued as

normal.

N =200.

Speculum inserted, vagina and

cervix cleaned and the cervix

grasped with a tenaculum. The

rigid, 2.7mm (3.5mm with

sheath) hysteroscope was then

inserted into the external

cervical canal and advanced into

the uterine cavity, using normal

saline as the distension medium.

The hysteroscopy then

continued as normal. N=200.

VAS 0-10cm, pain

rated immediately after

the procedure.

Mean and

standard

deviation and

median and range

reported.

62

Table 4.4 continued

Study

Participants

Intervention


Guida (78)

Pre- and post-menopausal



Exclusions: active genital tract

infection, cervical cancer,

heavy bleeding, severe

cardiovascular disease and

suspected pregnancy.

Rigid, 3.5mm minihysteroscope

introduced into the vagina, which

was distended with normal saline

so that the hysteroscope could be

steered into the external cervical os

and along the canal into the uterine

cavity. The hysteroscopy then

continued as normal. Endometrial

biopsies were taken when

indicated.

N =145.

Speculum inserted in to the

vagina and if necessary the

cervix was grasped with a

tenaculum. The 3.5mm

minihysteroscope was then

inserted into the external

cervical canal and advanced into

the uterine cavity, using normal

saline as the distension medium.

The hysteroscopy then

continued as normal.

Endometrial biopsies were taken

when indicated. N=145.

VAS 0-5cm marked

no, slight, tolerable,

severe and intolerable

pain. Asked to rate pain

at introduction of the

hysteroscope or

speculum into the

vagina, progression

through the cervical

canal, inspection of the

cavity and during the

endometrial biopsy.

Median and 95%

confidence

intervals.

Paschopoulos (79) Pre- and post-menopausal



Rigid, 3.5mm hysteroscope

introduced into the vagina. Holding

the labia minora together the vagina


and the hysteroscope steered into

the external cervical os and along

the canal into the uterine cavity.

The hysteroscopy then continued as

normal.

N=98.

Conventional hysteroscopy.

N=99.

VAS 0-10cm Divided into

groups scoring

<3, 3 – 5, >5.

63

Table 4.4 continued

Study

Participants

Intervention


Sagiv (33) Pre- and post-menopausal




vagina cleaned and the rigid,

3.7mm hysteroscope introduced

into the vagina which was then

distended with normal saline and

the hysteroscope steered into the

external cervical os and along the

canal into the uterine cavity. The


normal. Endometrial biopsies were

performed as necessary.

N=83.

Traditional hysteroscopy using a

rigid, 3.7mm hysteroscope and

normal saline, with an

intracervical injection of local

anaesthetic.

N=47.

VAS 0-10cm used to

score the pain

immediately and 15

minutes after the

hysteroscopy.

Mean and

standard

deviation

reported.

Sharma (32) Pre- and post-menopausal



Rigid, 3.5mm or 5mm hysteroscope

introduced into the vagina, which


so that the hysteroscope could be

steered into the external cervical os

and along the canal into the uterine

cavity. The hysteroscopy then

continued as normal. Endometrial

biopsies were taken when

indicated.

N =60.


vagina and the cervix was

grasped with forceps. Cervical

canal dilated if necessary and

local anaesthetic given if

requested by the patient. Either

a 3.5mm or 5mm rigid

hysteroscope was then inserted

into the external cervical canal

and advanced into the uterine

cavity, using normal saline as

the distension medium. The


normal. Endometrial biopsies

were taken when indicated

N=60.

VAS 0-10cm, rated at

insertion of

hysteroscope,

inspection of cavity,

insertion of speculum,

administration of local

anaesthetic,

endometrial biopsy, at

the end of the

procedure and 30mins

after the procedure.

Medians and

ranges.

64

Five papers reported data on the feasibility of the vaginoscopic procedure versus the

traditional hysteroscopy (31-33;77;78). Feasibility was assessed using the number of failed

procedures (i.e. failure to adequately visualise the uterine cavity) in each of the study arms.

Effect of vaginoscopy on pain

Meta-analysis of four studies showed that in the outpatient setting, the use of the vaginoscopic

approach to hysteroscopy was less painful than using the traditional technique with a vaginal

speculum (SMD -0.44, 95%CI -0.65 to -0.22, I2

= 58%) (Figure 4.14).

Figure 4.14 Forest plot showing the results of meta-analysis of studies that examine the

use of a vaginoscopic approach to outpatient hysteroscopy

One of these studies was considered ‘poor’ quality (79) as it only met one of the three quality

criteria. When this study was excluded from meta-analysis, the magnitude of reduction in pain

favouring vaginoscopy was not significantly altered and results were more homogenous

(SMD -0.52, 95%CI -0.71 to -0.33, I2

= 30%) (Figure 4.15). Neither of the two papers

included in the systematic review that could not be included in the meta-analysis (32;78)

reported any significant differences in the mean pain scores between the vaginoscopic and the

traditional hysteroscopy groups. However one of the studies (78) found that the 95%

65

confidence interval in the vaginoscopic group was significantly lower than in the traditional

hysteroscopy group (p<0.05).

Figure 4.15 Forest plot showing the results of meta-analysis of the high quality studies

that examine the use of a vaginoscopic approach to outpatient hysteroscopy

Feasibility

There was no significant difference in the number of failed procedures between the

vaginoscopic and traditional approaches to hysteroscopy p=0.38 (Figure 4.16).

Figure 4.16 Forest plot showing the results of meta-analysis of the studies that report

feasibility of the vaginoscopic approach to outpatient hysteroscopy

66

Results of the systematic review of the effect of distension media on pain



The literature search yielded 703 citations; twenty were retrieved for further evaluation.

Eleven studies were rejected because they did not meet the inclusion criteria (34;36-38;60;93-

98) and two were rejected as duplicates (64;93). The seven remaining studies were selected

as eligible for inclusion in the review (35;60;64;80-83); however it became apparent that two

of these papers may have reported the same study (81;99). The authors were contacted to

confirm this, however no answer was received. To prevent probable duplication of data only

the earliest published paper was used (81) resulting in six papers being eligible for the review

(Figure 4.17). The inter-rater reliability for the study selection was very good (kappa=0.85).

Details of the study populations, interventions, outcome assessment and data reporting are

shown in Table 4.5. The quality of the studies was poor with only one considered to have

adequate randomisation and concealment (81) this study was randomised by computer and so

it was assumed that it was also concealed.

67

Figure 4.17 Study selection process for systematic review of the effect of distension

media on pain during outpatient hysteroscopy

68

Table 4.5 Characteristics of the selected studies included in the systematic review of the effect of the distension medium used on pain

during outpatient hysteroscopy.

Study

Participants

Intervention

Comparison Outcome measures Data reported

Brusco (80)

Women attending for

hysteroscopy at an

artificial insemination and

sterility clinic.

External genitalia cleaned

with disinfectant. Outpatient

hysteroscopy performed with

a 4mm rigid hysteroscope and

a 7mm operative sheath by a

vaginoscopic approach.

Normal saline delivered by a

pressure bag (at 50mmHg)

used to distend the uterine

cavity. A paracervical block

was used when necessary.

N =45.

External genitalia cleaned with

disinfectant. Outpatient

hysteroscopy performed with a

4mm rigid hysteroscope and a

7mm operative sheath by a

traditional approach using a

speculum and tenaculum.

Carbon dioxide delivered by a

uterine insufflator at 40ml/min


cavity. A paracervical block

was used when necessary.

N =29.

Scale of 0-5 to score

pain during the

procedure. Also,

presence of shoulder

pain was recorded.

Operators graded the

quality of the image on a

scale from 0-5 (5= good

image quality).

Time from introduction

of the hysteroscope until

the removal at the end of

the procedure.

Pain and procedure

time reported as mean

and standard deviation.

Number and

percentage of women

experiencing shoulder

tip pain reported.

Image quality reported

as percentages.

Lavitola (81)

Infertile women



Vaginoscopic hysteroscopy

with a 30 degree continuous

flow hysteroscope, using

normal saline as the

distension medium introduced

by a pressure bag at 50-

120mmHg measured by a

manometer. An endometrial

biopsy was taken when

indicated.

N =97.

Vaginoscopic hysteroscopy

with a 30 degree continuous

flow hysteroscope, using

carbon dioxide as the

distension medium introduced

by uterine insufflator at a

pressure of 100mmHg. An

endometrial biopsy was taken

when indicated.

N =92.

VAS 0-10 used to score

pain during progression

of the scope through the

canal, during inspection

of the cavity, and the

intensity of any shoulder

pain.

The image quality was

scored by the operator

on a scale of 0-5

(5 = excellent).

The incidence of

vasovagal episodes.

Mean and standard

deviation for pain and

procedural time.

Number and

percentage of women

experiencing vasovagal

episodes and having

examinations with

mediocre image

quality.

69

Table 4.5 continued

Study

Participants

Intervention


Litta (82)


women undergoing

diagnostic outpatient

hysteroscopy.


vagina to visualise the cervix.

Rigid, 2.9mm, 30 degree

hysteroscope, introduced into

the cervical os and along the

canal into the uterine cavity

which was distended with

normal saline infused by a

100mmHg pressure bag.

Endometrial biopsies were

taken when indicated.

N =214.



Rigid, 2.9mm, 30 degree

hysteroscope, introduced into

the cervical os and along the

canal into the uterine cavity


carbon dioxide administered

by a hysterosufflator with a

pressure of 100mmHg and

flow rate of 40ml/min.


taken when indicated.

N =201.

VAS 0-10 completed

approximately 10


the procedure to rate the

pain experienced.

Procedure duration.

Presence of shoulder tip

pain.

Mean and standard

deviation for pain and

procedural time.

Number and

percentage of women


tip pain.

Nagele (83)


women undergoing


hysteroscopy.


vagina. Os probed and if

thought to be tight ,dilated

under local anaesthesia.

Rigid, 5.5mm , 30 degree

hysteroscopy with normal

saline infused by a 150-

250mmHg pressure bag.

Targeted endometrial biopsies

and minor surgical procedures

were performed when

indicated using a 7mm

operative sheath.

N =78.


vagina. Os probed and if

thought to be tight ,dilated

under local anaesthesia .Rigid,

5.5mm , 30 degree

hysteroscopy with carbon

dioxide distension media

administered by a

hysterosufflator at 100mmHg.

Targeted endometrial biopsies

and minor surgical procedures

were performed when

indicated using a 7mm

operative sheath.

N =79.

Abdominal pain and

shoulder tip pain ranked

on a scale of 0-4. (0= no

pain, , 4= very severe

pain).

Image quality was

graded on a scale of 0-4

(0= none, 1= good, 2 =

adequate, 3= poor, 4=

very poor).

Incidence of vasovagal

episodes.

Procedure duration.

Abdominal pain,

shoulder pain,

hysteroscopic vision

and procedural time

were all reported as

means and standard

deviations. The

number of women


pain and vasovagal

episodes were also

reported as numbers.

The number of poor or

very poor image

examinations was

reported.

70

PPI= Present pain intensity scale (verbal descriptors of pain ranked from 0-5 on a numeric scale). VAS= visual analogue scale

Table 4.5 continued

Study

Participants

Intervention


Paschopoulos (35)

Women admitted for total

abdominal hysterectomy

who agreed to undergo

outpatient diagnostic

hysteroscopy 12-24 hours

prior to surgery.

Exclusions: suspicion of

endometrial cancer

Hysteroscopy performed with

a 2.8mm, 30 degree, rigid

hysteroscope by a


Normal saline delivered by a

pressure bag (40-80mmHg)


cavity.

N =35.

Hysteroscopy performed with

a 2.8mm, 30 degree, rigid

hysteroscope by a


Carbon dioxide was delivered

by a microhysteroflator with a

maximum pressure of

200mmHg and a flow rate of

25ml/min to distend the uterine

cavity.

N =39.

Completed a

questionnaire after the

hysteroscopy which

asked patients to rate

shoulder pain and pelvic

pain on a 4 point scale

(0= none, 1= mild,

2=severe, 3= pain which

did not allow the

procedure to continue).

Number and

percentage of women

in each group selecting

each of the four

categories. Values

were allocated to the

categories i.e. 0, 1, 2, 3

and calculated the

mean and standard

deviation.

Shankar (64)


women undergoing


hysteroscopy for abnormal

uterine bleeding.

Exclusions: procedure not

feasible as unable to

visualise the cervix or

severe cervical stenosis.



Vulsellum applied. Cervix

dilated if necessary. Rigid,

5.5mm (with sheath), 30

degree hysteroscope,

introduced into the uterine

cavity which was distended

with normal saline infused by

a 150-250mmHg pressure

bag. Endometrial biopsies

were taken from all patients.

N= 100.

Another saline group had

lignocaine added to the

distension medium. This

group was not used in the

analysis. N=100



Vulsellum applied. Cervix

dilated if necessary. Rigid,

5.5mm (with sheath), 30

degree hysteroscope,

introduced uterine cavity


carbon dioxide, delivered by a

hysteron-insufflator at a

maximum pressure of

100mmHg with a variable flow

rate of up to 100ml/min.


taken from all patients.

N= 100.

Pelvic pain was scored

with VAS 0-10, and PPI.

Shoulder tip pain scored

with VAS 0-10.

The image quality was

ranked as very

satisfactory, satisfactory

or unsatisfactory.

Pelvic pain reported as

mean and standard

deviation.

Shoulder tip pain

reported as percentages

with 95% confidence

intervals. Percentages

were converted to

numbers.

Image quality reported

as number of

examinations falling

into each category.

71

Effect of distension media on pain

All six studies reported data for pain when comparing normal saline versus carbon dioxide as

the distension media for the procedure. Five studies used scales to collect the data (64;80-83)

and reported them as means and standard deviations. One study reported the number of people

selecting each of four numbered categories (0= none, 1= mild, 2=severe, 3= pain which did

not allow the procedure to continue) (35). The category number was used as a value and the

mean and standard deviation were calculated (90) for the pain experienced during the

procedure. Meta-analysis showed that there was no significant difference in pain scores when

carbon dioxide or normal saline were used as the distension medium for outpatient

hysteroscopy (SMD= -0.05, 95% CI -0.17 to 0.07, I² = 92%) (Figure 4.18).

Figure 4.18 Forest plot showing the effect of distension media on procedural pain during


Effect of distension media on shoulder tip pain

Shoulder tip pain was reported by all six studies. One study used a visual analogue scale to

assess the severity of the pain and reported a mean and standard deviation (81). The

remaining five studies (35;64;80;82;83) reported the number of events or the percentage of

women that experienced shoulder tip pain. If the number of events were not reported (64;82)

72

the value was calculated from the percentage. Meta-analysis of five studies found that

shoulder tip pain was significantly reduced when using normal saline as compared to carbon

dioxide as the distension medium (OR = 0.19, 95% CI 0.09 to 0.40, I²= 41%) (Figure 4.19).

Figure 4.19 Forest plot showing the effect of distension media on shoulder tip pain

during hysteroscopy

Effect on vasovagal episodes

Vasovagal reactions were specifically reported by four studies (35;81-83), one of which

reported no vasovagal episodes (82). The two studies that didn’t report vasovagal episodes did

report symptoms that may be attributed to vasovagal episodes (nausea, dizziness,

hypotension) (64;80) but did not specifically state them to be vasovagal reactions. The

number of events from the three studies (35;81;83) that stated the patients had suffered

vasovagal episodes were used for meta-analysis and it was found that there were significantly

fewer vasovagal episodes when using normal saline as the distension medium (OR = 0.31,

95% CI 0.12 to 0.82, I²= 0%) (Figure 4.20).

73

Figure 4.20 Forest plot showing the effect of distension media on vasovagal episodes


Effect on duration of the procedure

Procedural time was reported by four studies (80-83) as means and standard deviations. Meta-

analysis found that outpatient hysteroscopy using normal saline was significantly shorter than

when using carbon dioxide (SMD=-1.32, 95% CI -1.48 to -1.17, I² =98%) (Figure 4.21).

Figure 4.21 Forest plot showing the effect of distension media on procedural time for


Effect of distension media on image quality

Image quality was reported by four studies (64;80;81;83). Three studies used scales with

categories (e.g. 0= no view, 1= poor view etc.) (80;81;83) and asked the operators to select

the appropriate number. One study used this data to calculate a mean and standard deviation

74

(83). The remaining two studies (80;81) reported the number of operators selecting from

certain categories but did not give data for all of the categories. The final study (64) used

unnumbered categories (very satisfactory, satisfactory, unsatisfactory) and asked the operators

to select an appropriate one. Data for the ‘very satisfactory’ category were reported in the text

and for the ‘unsatisfactory’ category in a table, thus allowing calculation of the number of

operators selecting the ‘satisfactory’ category. Data from the studies could not be meta-

analysed as they were not adequately reported and outcome assessments were not comparable.

Three studies reported no significant difference in image quality between carbon dioxide and

normal saline (80;81;83), however one of these studies (83) reported changing the distension

media from carbon dioxide to normal saline in 8 (10.1%) patients. One study found a

statistically significant increased risk of unsatisfactory view with the use of carbon dioxide

(Relative risk (RR) =4.75, 95% CI 1.61, 16.4) attributing it to bubbles, and bleeding (64). Of

the 19 patients who had an unsatisfactory view at hysteroscopy using carbon dioxide, 17 were

changed to normal saline and an improved view reported in 11 (64.7%).

75

Results of the systematic review of the effect on pain of the type of

hysteroscope used for outpatient hysteroscopy


From 214 citations, 212 were rejected on methodological grounds or because they were

duplicates. That left just two studies (84;85) which were selected for review (Figure 4.22).

Figure 4.22 Study selection process for systematic review of the effect on pain of the type

of hysteroscope used for outpatient hysteroscopy

76

Both selected studies were randomised controlled trials that compared flexible with rigid

hysteroscopes, using normal saline as the distension medium. The inter-rater reliability for the

study selection was very good (kappa=0.84). Pain scores were assessed on 10cm visual

analogue scales in both studies. Both studies assessed pain immediately after the procedure

(84;85) and one assessed pain during the procedure (85). The study characteristics and results

are shown in Table 4.6.

Pain during the procedure was found to be significantly less when the flexible hysteroscope

was used (flexible median pain score 1.2, rigid median pain scores 3.1, p = 0.0001 (85)).

Immediately after the hysteroscopy the studies reported different results, with one finding a

significantly lower pain score when a flexible hysteroscope was used (flexible pain scores

mean 1.8, median 1.2 versus rigid pain scores mean 4.0, median 3.6, p= 0.0001) (84) and the

second finding no difference between pain scores (median 0 in flexible and rigid groups) (85).

No vasovagal episodes occurred in the single study reporting this outcome (84). Image quality

was assessed by the hysteroscopists and reported in both studies. One study (84) reported that

the view was good to excellent in all examinations whereas the other study (85) reported a

significantly better view with the use of the rigid hysteroscope (p< 0.001). There were no

failures in the rigid hysteroscopy groups, however one of the studies (85) reported five

failures in the flexible hysteroscopy group. This is reported as ‘(5/40) 12.5%’ in the paper,

however there were 70 rather than 40 patients in the group which results in a 7.2% failure

rate. One of the studies found use of a rigid hysteroscope to be significantly quicker than

using a flexible scope (p = 0.003) (85) in contrast to the second paper which reported similar

procedure times in both groups (flexible 5.9 minutes, rigid 6.1 minutes) (84).

77

Table 4.6 Overview of the studies included in the systematic review of the effect on pain of the type of hysteroscope used for outpatient

hysteroscopy including interventions and results

Baxter (84)

Unfried (85)

Design

Randomised controlled trial Randomised controlled trial

Randomisation Computer generated random number sequence Computer generated random permutation list with 36

blocks of length four

Blinding Single blinded Not reported

Follow-up 100% 100%

No. in flexible group 40 70

No. in rigid group 43 72

Intervention After a vaginal examination to assess uterine size and

position, a Cusco’s speculum was inserted and the

cervix was visualized and cleaned. When necessary, the

anterior lip of the cervix was grasped gently with a

tenaculum. Occasionally a local anaesthetic block was

used. Hysteroscopy was then performed with either a

3mm flexible hysteroscope (Olympus Keymed,

Southend-on Sea, UK) or a 3.3mm rigid scope,

including sheath (Storz, Slough, UK). The distension

medium was warmed normal saline at a pressure of up

to 100 mmHg.

The procedure was performed according to a standard

protocol which is not described in the paper. Lignocaine

spray was applied to the cervix before the hysteroscopy

and occasionally an additional local anaesthetic block

was used. Hysteroscopy was then performed with either

a 3.6mm flexible hysteroscope with 100° bending

section (Olympus HYF type P, Hamburg, Germany) or

a 3.7mm rigid hysteroscope with a 30° fore oblique lens

(WISAP, Munich, Germany). The distension medium

was normal saline at a pressure of 120-200 mmHg.

78

Table 4.6 continued

Baxter (84)

Unfried (85)

Method of scoring pain Analogue pain scale measuring 10 cm long, with a score

of 0 at one end corresponding to having ‘no pain at all’

and 10 at the other being ‘the worst pain they have ever

had’. Completed immediately and 30 minutes after the

hysteroscopy.

A 10cm visual analogue scale completed before the test,

at insertion of the hysteroscope, during and immediately

after the hysteroscopy and at endometrial biopsy.

Pain score in flexible group During hysteroscopy- not reported

Immediately after- mean 1.8, median 1.2

During hysteroscopy- Median 1.2

Immediately after – 0

Pain score in rigid group During hysteroscopy- not reported

Immediately after- mean 4.0, median 3.6

During hysteroscopy- Median 3.1

Immediately after – 0

P value of pain score difference During hysteroscopy- not reported

Immediately after p= 0.0001

During hysteroscopy- p<0.001

Immediately after- p= NS

Vasovagal episodes in flexible group 0 Not reported

Vasovagal episodes in rigid group 0 Not reported

Failures in flexible group 0 5 (7.2%) –reported as 12.5% in the paper.

Failures in rigid group 0 0

Quality of image in flexible group Excellent to good 66% excellent to good

Quality of image in rigid group Excellent to good 100% excellent to good

79

CHAPTER 5

DISCUSSION REGARDING THE RESULTS OF THE

SYSTEMATIC REVIEWS AND META-ANALYSES WHICH

INVESTIGATE PAIN REDUCTION IN OUTPATIENT

HYSTEROSCOPY

Principal findings of the reviews

The systematic review of local anaesthetic found that pain is reduced during outpatient

hysteroscopy with paracervical and intracervical injections of anaesthetic but not with

transcervical and topical application. Paracervical injection appears to be the most effective

method of administering local anaesthetic for the procedure. Local anaesthetic did not

significantly reduce the incidence of vasovagal attacks during outpatient hysteroscopy but

there was a beneficial trend.

The analgesia systematic review found that opiates may reduce the pain of outpatient

hysteroscopy during the procedure but not without significant side effects, although some of

the symptoms experienced (nausea, vomiting, dizziness) could be attributed to the procedure.

In one study the side effects attributed to buprenorphine caused high rates of patient

dissatisfaction, demonstrating the importance of considering whether the benefit of pain relief

outweighs medication side effects. A study, examining mefenamic acid found a significant

reduction in post-operative pain at 30 and 60 minutes.

80

Conscious sedation with midazolam was found to be as effective as intracervical local

anaesthetic for reducing pain during outpatient hysteroscopy. Sedation requires regular

monitoring and may be associated with inadvertent overdose, requiring reversal.

Hysteroscopists will be experienced in administering an intracervical block but will have far

less experience in the use of conscious sedation and are unlikely to feel confident to manage

the sedation without an anaesthetist or a second clinician. Additionally, ambulatory

gynaecologists who perform colposcopy are already familiar with techniques for anesthetising

the cervix when taking biopsies. Thus to ensure patient safety and reduce the number of

trained staff required, alternatives to conscious sedation should be used for outpatient

hysteroscopy.

The studies used in the systematic review of cervical preparation examined different drug

doses and administration times and assessed pain at different time points. Similarly,

effectiveness of cervical priming was assessed in a variety of ways. This heterogeneity made

meta-analysis unfeasible and interpretation of the results problematic. However the following

conclusions were drawn from the results. Mifepristone given before hysteroscopy does not

have any significant effect on the pain experienced during the procedure or dilatation of the

cervix in premenopausal women. Cervical priming with prostaglandins is beneficial in

reducing pain when dilating the cervix beyond 6mm but this is likely to be greater in

postmenopausal women. Pain on inserting the hysteroscope and during the hysteroscopy is

reduced when using large diameter hysteroscopic systems (4mm endoscopes with 5.5mm

sheath) but miniaturisation of the instruments used in the outpatient setting means that there is

rarely any need to dilate the cervix above six millimetres thus these findings are, in the most,

clinically irrelevant. It was demonstrated that prostaglandins do not have significant adverse

81

effects when compared to placebo and do not increase the number of failed procedures,

however the negligible benefit means that the routine use of cervical priming prior to

outpatient hysteroscopy cannot be recommend.

The review of vaginoscopy for outpatient hysteroscopy found that the vaginoscopic approach

to hysteroscopy significantly reduces the pain experienced by patients during the procedure.

The vaginoscopic technique is successful as demonstrated by the majority of hysteroscopies

being completed successfully (83-98%) and there being no significant difference in the

number of failed procedures between groups of patients undergoing a vaginoscopic

hysteroscopy and a traditional outpatient hysteroscopy using a vaginal speculum. However

whilst a reduction in pain is clearly advantageous in outpatient procedures to optimise

acceptability to patients, the review did not demonstrate any improvement in procedural

feasibility (i.e. successful completion of hysteroscopy) as a consequence of minimising

discomfort.

There is no significant difference in pain experienced between carbon dioxide and normal

saline when used as the distension medium for outpatient hysteroscopy. However, meta-

analysis found that vasovagal episodes, shoulder tip pain and procedural time were all

significantly reduced when using normal saline. Image quality may be better with normal

saline as it causes a lavage and so prevents blood and bubbles obscuring the view.

Flexible hysteroscopy causes less pain than rigid hysteroscopy in the outpatient setting. Rigid

hysteroscopy may provide a better image, resulting in fewer failed procedures and be quicker

to perform although these results were not consistently reported. Mean pain scores were low

82

with both types of hysteroscope (≤4) and so without further qualitative assessment it is

difficult to assess the clinical relevance of the lower pain score associated with flexible

hysteroscopy, especially when procedure times may be more prolonged (85) and more likely

to fail (85) with a non-rigid endoscope. Whilst the current review did not evaluate test

accuracy, the superior image quality with rigid hysteroscopy reported in one study (85) may

further offset any clinical advantage of a less painful procedure using a flexible instrument.

Strengths and limitations of the reviews

Many aspects of the reviews support their results being valid. Firstly, clinically focused

questions were formulated. Comprehensive searches were then performed which

encompassed multiple online databases as well as searching of the reference sections of

relevant studies. The searches were not restricted to the English language and broad search

terms were used to avoid making the question too specific to be adequately sensitive.

Unpublished data were not sought and therefore there is a risk of publication bias. Data were

restricted by the study design to randomised controlled trials to minimise selection bias.

Excepting the review of local anaesthesia, all of the reviews contained small numbers of

studies which limits the conclusions. However, although there was paucity of studies the

reviews did contain the totality of published, randomised data. Meta-analysis was not always

possible making it difficult to draw conclusions for the examined elements of hysteroscopy in

some of the reviews.

For the systematic review of local anaesthetic the studies were sub-grouped according to

quality to examine for the overall heterogeneity. Heterogeneity was reduced in the high

83

quality sub-group, but I2

was still 83%. Meta-analysis of the high quality studies found a

significant reduction in pain with the use of local anaesthetic, a finding which was consistent

with and thus supportive of the overall findings. Intracervical injection of local anaesthetic

was associated with a significant reduction in pain during outpatient hysteroscopy, but the

strength of this finding is limited. This is because a sensitivity analysis excluding a study in

which categorical data had been transformed, demonstrated no beneficial effect of

intracervical injection, although this finding was associated with increased heterogeneity.

Reasons for heterogeneity could not be examined in the method of administration sub-groups

due to the paucity of studies.

Out-patient hysteroscopy is a multi-faceted procedure and so there are many factors that

contribute towards pain. These can be categorised into patient factors (e.g. menopausal status,

reason for hysteroscopy) and procedural factors (e.g. the type of distension media, use of a

speculum, use of a rigid or flexible hysteroscope). The small number of studies in the review

of local anaesthesia meant that there were not enough data to sub-group the patients according

to menopausal status and indication for hysteroscopy. Confounding due to procedural factors

should be eradicated on the basis of the study designs being restricted to randomised

controlled trials in this review.

There were further limitations in the study of the cervical preparation, mainly due to the

heterogeneity of the studies. Although the majority of studies gave 400µg misoprostol (71;74-

76) the administration times varied between the studies from four, to twenty-four hours before

the procedure. One study used mifepristone as a priming agent (72) and so was incomparable

with the other five studies The studies varied in the timing of pain assessment with some

84

assessing pain during cervical dilatation whilst others assessed pain during or after the

procedure. Similarly, cervical dilatation was assessed in a non-uniform manner, some studies

assessing the force needed to dilate and others assessing change in dilatation from baseline or

dilatation pre-procedure. The study populations differed so that conclusions regarding

menopausal status and the use of cervical priming agents were limited.

Only the abstract was available for one of the studies included in the vaginoscopy systematic

review (79) and only limited information about the study method was available hence it was

graded as low quality. However, the five remaining RCT’s were all considered to be of high

quality. All of the studies used a tenaculum in the ‘traditional’ hysteroscopy arm and thus it

was not assessed whether there was a significant pain reduction when there is no cervical

instrumentation. Whilst all studies included in the review compared vaginoscopy with a

traditional approach to hysteroscopy utilising a vaginal speculum, there were some minor

procedural differences between the studies (e.g. use of intracervical local anaesthetic (33),

varying distension media between the arms of the study (77)), however heterogeneity was low

in all analyses, showing consistency of findings across studies in favour of vaginoscopy.

Effectiveness of the procedure in sub-groups of patients (e.g. pre- versus post-menopausal,

nulliparous versus parous) could not be assessed as all of the studies included mixed

populations. There were no studies which used flexible hysteroscopes for vaginoscopy or that

compared vaginoscopy to flexible hysteroscopy.

There was considerable heterogeneity (I2

=92%) in the meta-analysis of procedural pain when

comparing distension media for outpatient hysteroscopy. Because the random effects model

gives more weight to small studies, one small study (80) with an outlying result was excluded

85

to explore heterogeneity. As well as being small this study used a vaginoscopic approach

(without a speculum and tenaculum) in the normal saline group and a traditional approach

with a speculum in the carbon dioxide group and so had more than one variable. Analysis

without this study reduced the heterogeneity to 82% but still produced a non-significant result

overall (SMD = 0.06, 95% CI -0.24 to 0.37). No causes for heterogeneity were identified but

may be attributable to the small number of studies used for the meta-analysis.

Heterogeneity was also high in the meta-analysis of procedural time (I2

=98%), again this was

difficult to explain. All of the studies in this analysis found individual significant results in

favour of the use of normal saline, which were in keeping with the meta-analysis result. In

contrast, heterogeneity was low in the meta-analyses exploring vasovagal episodes and

shoulder tip pain. A weakness of this distension media review is that it only contains six

studies, five of which were assessed to be of low quality due to inadequate randomisation and

concealment and two of which have small populations. However, as already mentioned, these

data reflect the totality of the published trials in this area. Although statistically significant

results were found it was not possible to assess whether they were clinically significant.

Comparison with other studies

The systematic reviews that examine local anaesthetic, vaginoscopy, cervical preparation and

distension media have been published in peer reviewed journals (100-103) since this thesis

was undertaken. Prior to that there were no published systematic reviews that addressed

methods to relieve pain during outpatient or office hysteroscopy. The conclusions of these

reviews were unique in having been drawn from the available primary studies.

86

A Cochrane review had assessed the use of paracervical injection for a variety of obstetric and

gynaecological procedures including hysteroscopy, endometrial biopsy, fractional curettage,

vacuum aspiration, suction termination of pregnancy or evacuation of retained products of

conception and bimanual removal of retained placenta (14). Only three studies involving

hysteroscopy were included in the review and only two in the meta-analysis so their

conclusion that the use of paracervical injection does not reduce the pain of ‘uterine

intervention’ cannot be applied specifically to hysteroscopy. The meta-analysis of

paracervical injection in this systematic review contains five studies assessing paracervical

anaesthesia in hysteroscopy and so has greater power, adding weight to the findings.

Moreover, intracervical and topical administration of local anaesthesia is more commonly

practised in ambulatory hysteroscopy than paracervical approaches (6). The systematic review

of local anaesthesia described in this dissertation assesses all routes of local anaesthetic

administration thereby providing relevant guidance to clinicians for one of the most

commonly practised interventions in gynaecology. Similarly, a Cochrane review published in

2012 examined analgesics and local anaesthetics (104) but did not separate the anaesthetics

according to method of administration. This review agreed that local anaesthetics were

successful at reducing pain during outpatient hysteroscopy but by failing to look at method of

administration it is difficult to make clinical inferences and decide how to administer the

anaesthetic for maximum effect. In fact, the review of local anaesthetic in this thesis was able

to demonstrate that the topical methods of administration do not have any benefit in pain

reduction and thus provide clinically useful information.

87

The Cochrane review of local anaesthetics and analgesia (104) concluded that there was

insufficient evidence for the use of analgesics prior to outpatient hysteroscopy which is in

keeping with the review conducted for this thesis. A review that assessed the use of analgesia

for hysterosalpingogram (HSG) found that when using analgesia versus placebo there was no

benefit of using the analgesia during the procedure, however there was a benefit 30 minutes

post-procedure (105). This is in keeping with the results of the outpatient analgesia review

and although it refers to a different procedure, HSG is an outpatient test which also involves

distension of the uterus in a similar way to hysteroscopy.

A systematic review of the use of paracervical local anaesthesia for cervical dilatation and

uterine intervention (14) assessed conscious sedation versus cervical block for hysteroscopy

and bimanual removal of placenta and found there to be no difference in pain between the

sedated patients and those receiving local anaesthetic. This is in keeping with the findings of

the only study retrieved in the review of conscious sedation.

A review that examined the use of misoprostol prior to hysteroscopy (studies examining

procedures under general anaesthetic and in the office) (24) found that misoprostol given to

premenopausal women reduced the need for cervical dilatation as well as the incidence of

cervical laceration. However, vaginal bleeding, abdominal cramping and fever were

significantly increased in the misoprostol groups. No significant difference was found

between the control and intervention groups in postmenopausal women. These results conflict

with the findings of the review of cervical preparation for outpatient hysteroscopy, mainly

because the majority of the studies examined were assessing the use of misoprostol for

inpatient hysteroscopy which routinely requires cervical dilatation because of the use of larger

88

instruments. Secondly as most of the patients were anaesthetised the review was unable to

assess pain during the procedure. The systematic review of cervical dilatation for outpatient

hysteroscopy was only able to meta-analyse a small number of studies to assess the adverse

effect of prostaglandins and thus the result may underestimate their true incidence.

Two recent systematic reviews have examined the use of misoprostol given prior to

hysteroscopy (106;107). The first examined the benefits and harms of misoprostol given prior

to operative hysteroscopy with hysteroscopes of greater than 9mm in diameter. These

procedures were not carried out in the outpatient setting but despite the use of much larger

hysteroscopes, there was still no benefit of using misoprostol for dilatation of the cervix or in

reducing complications. Similarly to the review by Crane et al. (24), side effects were

significantly increased in the group of women who had taken misoprostol. This is not in

keeping with the results found in this thesis as discussed above. The second review (107)

investigated the feasibility of using misoprostol to facilitate the passage of the hysteroscope

and the side effects associated with the medication. Pain during the hysteroscopy was not

assessed. Studies that included inpatient and outpatient procedures were included, resulting in

17 studies being eligible. However, this will have caused heterogeneity as the size of

hysteroscopes used will vary. The authors do not discuss this. This review found that

misoprostol significantly reduced the need for cervical dilatation in the total population but

when the women were analysed according to menopausal status there was no significant

difference for either pre or postmenopausal women. The difference found for women overall

may have been driven by the studies which used large hysteroscopes and although the

differences were statistically significant they may not be clinically significant. Once again, the

presence of side effects was significantly increased in the misoprostol group. The authors

89

concluded that there was not enough evidence to recommend routine use of misoprostol prior

to hysteroscopy.

A large case series of five thousand patients found that carbon dioxide and normal saline were

both feasible distending media, with no difference in success rate but that there were

significantly more vasovagal episodes and shoulder tip pain when carbon dioxide was used

(93). This is in keeping with the findings of the distension media review.

Clinical implications of the reviews

Injectable, preferably paracervical, administration of local anaesthetic should be used for out-

patient hysteroscopy to reduce the amount of pain that women experience. Topical application

of local anaesthetic does not reduce the pain of the hysteroscopy but should be used when

applying a tenaculum to the cervix. Only one study examined the use of local anaesthetic for

operative hysteroscopy (60) and therefore the results do not adequately address the benefit of

local anaesthetic in this variation of the procedure.

Although these conclusions show a benefit of using local anaesthetic, data reviewing the

harms could not be reviewed. This is because harms were not explicitly reported by the

majority of studies except when referring to symptoms caused by vagal stimulation

(hypotension, bradycardia, nausea, vomiting etc.) although it is possible that similar

symptoms may arise from intravasation of injected local anaesthetic. Therefore side effects

resulting from the use of local anaesthetic for outpatient hysteroscopy are likely to be

underestimated. For example, the time taken to perform the block prolongs the procedure and

90

the pain scores may not take into account the pain experienced during injection of local

anaesthetic, in itself a painful procedure. In fact it was found in one study that a vaginoscopic

approach to hysteroscopy was significantly less painful than having the procedure done

traditionally using a vaginal speculum and a local anaesthetic block (33).

Good pain control is important to ensure an efficient outpatient ‘ambulatory’ hysteroscopy

service with rapid discharge and high levels of patient satisfaction (6;108). When prescribing

analgesia non-steroidal anti-inflammatories should be preferred to opioids as they have fewer

side effects but should be avoided in certain patient groups (asthmatics, history of peptic

ulceration, allergy).

For investigation of the gastrointestinal tract, which cannot be anaesthetised and may require

quashing of the gag reflex, conscious sedation is an invaluable resource. However, conscious

sedation requires an experienced clinician to monitor the degree of sedation and possibly

reverse the medication (15) and this is an unfamiliar entity to the gynaecologist. As there is no

benefit of conscious sedation over paracervical block (a technique familiar to the

hysteroscopist and administrable by them) it would seem that conscious sedation should be

avoided for outpatient hysteroscopy.

The review of cervical preparation suggests that routine use of prostaglandins prior to

outpatient hysteroscopy would not provide a clinically significant change in dilatation or pain

and although no significant adverse effects were found, other systematic reviews of their use

suggest that this may have been underestimated. A further clinical consideration is the added

cost of using cervical preparations. As well as the cost of the drugs there may be increased

91

costs related to the added intervention, such as extra clinic appointments and management of

adverse effects. There may need to be a facility for administering the vaginal drugs, especially

in elderly women. Whether they are admitted for the prostaglandin and stay until the

procedure or they go home between, there will be significant inconvenience for the patient

and a possible impact on hospital beds. The final consideration is patient acceptability.

Patients may prefer to have a short period of discomfort during the procedure, rather than

abdominal cramping and vaginal bleeding for hours beforehand.

The findings that vaginoscopy is associated with significantly reduced procedural pain and

comparable feasibility, compared with traditional speculum assisted and cervical

instrumentation approaches to office hysteroscopy have clinical implications for clinical

practice. Outpatient hysteroscopy is a common diagnostic and increasingly therapeutic

intervention (e.g. polypectomy, sterilisation) and vaginoscopy is not widely taught or

practised (6). The findings of this review however, strongly support vaginoscopic outpatient

hysteroscopy as the technique of choice. Although it was not examined by this review, it is

proposed that the vaginoscopic approach allows greater intracavity manoeuvrability of the

hysteroscope as there is no restriction on external movement caused by the presence of a

vaginal speculum. This is particularly advantageous in certain patient groups (e.g. restricted

hip abduction / flexion, inability to lie supine due to respiratory disease, obesity, enlarged or

acutely flexed uteri) and when access to the tubal ostia is required, for example, during

hysteroscopic sterilisation (109). In addition, nulliparous women, those who have not been

sexually active, have atrophy of the genital tract or who find intimate examinations distressing

are likely to find the vaginoscopic approach more acceptable as it obviates the need for

stretching the vagina by a bivalve speculum. Traditional hysteroscopy, with a vaginal

92

speculum and possibly cervical instrumentation, will still be necessary in the minority of

cases where visualisation of the cervical canal is difficult (e.g. deficient intravaginal cervix,

previous cone biopsy) or cervical stenosis is encountered. A vaginal speculum is also

required when a global endometrial biopsy is indicated although there are new developments

in global hysteroscopic endometrial sampling (110). Adopting a first-line approach of

vaginoscopy allows visualisation of the uterine cavity with a subsequent recourse to either

directed hysteroscopic biopsy or global ‘blind’ office sampling of the endometrium.

Normal saline and carbon dioxide are both suitable for outpatient hysteroscopy, as pain

experienced during the diagnostic procedure is comparable between media. However, normal

saline is the distension medium of choice in the office setting because it is associated with

significantly fewer vasovagal episodes and incidences of shoulder tip pain. Furthermore,

distension of the uterine cavity with normal saline as opposed to carbon dioxide results in a

quicker hysteroscopic procedure, which is especially advantageous in conscious patients

undergoing an invasive test. Although image quality could not adequately be assessed two of

the studies reported changing from carbon dioxide to normal saline when the image was poor

or obscured by blood (64;83) thus suggesting that normal saline is a more practical medium

and can cause an improvement in view over carbon dioxide. A further consideration is the

specialist equipment that is required to use carbon dioxide to distend the uterus. A

hysterosufflator is required to control insufflation of carbon dioxide into the uterus however

normal saline can be administered simply with a giving set and gravity or a pressure bag,

although this can result in leakage of fluid and a more ‘messy’ procedure, compared with

carbon dioxide. This review did not examine cost-effectiveness of the distension media but

the increased duration of the procedure and capital outlay for specialist equipment associated

93

with the use of carbon dioxide, for no apparent benefit over normal saline, may not support

the routine use of carbon dioxide in an outpatient setting from an economic perspective. A

final, important consideration when choosing distension medium refers to the increasing use

of therapeutic procedures in outpatient hysteroscopy. These procedures include the use of

bipolar electrosurgery (6;111) for removing fibroids, polyps, synechiae (112) and septae

(113;114) which require a conducting liquid medium (i.e. normal saline). Moreover, other

office interventional procedures using mechanical equipment (115) or where there is a

likelihood of endometrial fragmentation, necessitate the use of a liquid distension medium to

maintain visualisation by clearing both blood and debris. Thus performing a diagnostic

hysteroscopy with normal saline will negate the need to change distending medium should a

subsequent operative procedure be required (i.e. the ‘see and treat’ ethos).

The review of hysteroscope type was not able to address the differences between the types of

scope with regards to cost and maintenance. It has been reported that flexible hysteroscopes

are approximately twice as expensive to purchase (40;116) as rigid hysteroscopes and more

expensive to sterilise and maintain (116). The clinical implications of this review are that

when purchasing hysteroscopes, clinicians should consider the advantage of flexible

hysteroscopy as regards to reduced procedural pain against the potential disadvantages of

procedure duration, feasibility and cost.

Unanswered questions and future research

These reviews have implications for future research. They highlight the need for large trials

comparing how the different hysteroscopic techniques (e.g. vaginoscopy versus a traditional

94

hysteroscopy with or without local anaesthesia, type of distension media, use of a flexible or

rigid scope, cervical preparation etc.) affect pain, feasibility and the incidence of vasovagal

episodes during outpatient hysteroscopy. Such trials should explicitly define and standardise

the procedure and systematically examine acceptability and quality of life, in addition to

alleviation of pain. These qualitative outcomes can then be correlated with pain scores to see

if any reported reduction in pain during outpatient hysteroscopy is actually clinically

meaningful. Is the pain that women feel during a hysteroscopy enough to warrant the use of

anaesthetic? It may cause a significant reduction in the amount of pain experienced, however

in the majority of studies the mean pain scores felt in the intervention and control groups

showed little variation and tended to be low anyway. Then it must be considered that case

selection is important when assessing who will benefit the most from the use of local

anaesthetic. Parous, pre-menopausal women are less likely to benefit as it is presumed that

they have less narrowing of the cervical canal, whereas nulliparous post-menopausal women

who will almost certainly have a degree of cervical stenosis may benefit greatly. The results

from these reviews cannot quantify this benefit. It may be that altering different aspects of the

procedure (e.g. vaginoscopy, warming liquid distension media, distension pressures etc.) can

reduce the pain significantly without having to use local anaesthetic injection. There are

obvious cost implications (for equipment and medication, as well as reducing the number of

people seen in clinic) of using local anaesthetic for every patient who undergoes an outpatient

hysteroscopy.

Research trials should also evaluate administration of local anaesthetic, analgesics and

hysteroscopic technique in operative outpatient hysteroscopic surgery, which is becoming

increasingly prevalent with technological advances in endoscopic instrumentation

95

(6;117;118). The timing of administration needs to be assessed as well as the different

available drugs and doses and the incidence of adverse effects. It would be prudent to assess

the benefit of simple analgesics regimens (e.g. paracetamol or ibuprofen) self-administered by

the patient prior to attending for their examination. The use of ‘rescue analgesia’ once the

patient has left the hospital may be indicative of the medium term benefit of analgesics and

this information could also be collected along with qualitative data referring to the patient

experience and rates of satisfaction. Additionally, the role of simply administered, short

acting inhalational sedation (e.g. EntonoxTM

- nitrous oxide with oxygen) should be examined

with randomised studies, particularly for therapeutic outpatient hysteroscopic procedures.

It is during blind dilatation of the cervix that perforation of the uterus is most likely to occur.

There are subgroups of women (previous caesarean section, nulliparous, postmenopausal,

previous loop biopsy) who more often need dilatation of the cervix to facilitate passage of the

hysteroscope. A well designed clinical trial should concentrate on these subgroups to assess

whether the use of prostaglandins provides a clinically significant reduction in pain, reduces

the need for dilatation and incidence of genital tract trauma and whether these benefits

outweigh the potential harms (bleeding, laceration, abdominal cramping, fever). As well as

focussing on certain populations the study should use miniature hysteroscopes to make the

results clinically relevant to modern practice. If positive results were found, further research

would be needed to determine optimum dose and timing of administration of the

prostaglandins.

Large trials are needed to compare vaginoscopy to traditional hysteroscopy using both rigid

and flexible hysteroscopes. Such trials should explicitly define and standardise the procedure

96

and systematically examine acceptability and quality of life, in addition to alleviation of pain,

the risk of ascending infection and the incidence of vasovagal episodes. In addition, studies

need to look a priori at patient factors, such as parity, menopausal status, previous cervical

biopsy or caesarean section and body mass index to identify patient sub-groups in whom the

vaginoscopic procedure is less likely to succeed.

An economic analysis should be performed to analyse whether there is any true discrepancy

regarding cost when different distension media or flexible or rigid hysteroscopes are used for

outpatient hysteroscopy. Personal experience is that the majority of clinicians within the UK

are using normal saline and rigid hysteroscopes and it would be interesting to examine the

reasons for this with qualitative research.

Summary

The ideal outpatient hysteroscopy would be pain free. However this is impossible and so the

utmost should be done to minimise pain. The seven systematic reviews conducted suggest that

in order to optimise success and minimise pain, women undergoing outpatient hysteroscopy

should take simple analgesia a short time prior to the procedure and the hysteroscopist should

adopt a vaginoscopic approach using a small diameter, rigid hysteroscope and normal saline

as the distension medium. If dilatation of the cervix is required, this should be done under a

paracervical block. The use of cervical preparation and conscious sedation should not be

routinely used.

97

CHAPTER 6

EVIDENCE QUALITY IN CLINICAL GUIDELINES: A

COMPARISON OF TWO METHODS

In the 21st century we no longer practice medicine based on our own experience and

prejudices but try to find real evidence that our practice is not only beneficial but that it does

not harm and is cost-effective. This process is called evidence based medicine and should be

the foundation of modern medical care.

Guidelines and protocols are developed using evidence relevant to their topics and this should

be critically appraised during development. The Royal College of Obstetricians and

Gynaecologists use the SIGN methodology (119;120) for their clinical, Green-top guidelines,

which are used to inform practice in Obstetrics and Gynaecology throughout the UK and

internationally.

‘Quality of evidence’ has been defined as ‘reflecting the extent to which confidence in the

estimate of an effect is adequate to support recommendations’ (121). Guidelines report the

quality of the evidence used when formulating recommendations so that clinicians can decide

how reliable the recommendation might be. Unfortunately, different organisations use

different methods of grading the quality of evidence which can make interpretation difficult

(122), particularly if multiple guidelines are being considered. In view of this, the Grading of

Recommendations Assessment, Development and Evaluation (GRADE) working group have

98

developed the GRADE system which is intended as a common, sensible and transparent

approach to grading the quality of evidence (123) for use in clinical guidelines. The GRADE

format (122) claims to make it easier for users to assess the judgments behind

recommendations by tabulating the evidence.

Using the systematic reviews of outpatient hysteroscopy presented in this thesis, a guideline

was written for the British Society of Gynaecological Endoscopists (BSGE) in the format of

an RCOG Green-top guideline, entitled ‘Best Practice in Outpatient Hysteroscopy’. This

guideline has since been published by the RCOG as a formal Green-top guideline (124) and is

available to clinicians in the UK and abroad (Appendix 4). To investigate how the evidence

quality would have been graded if the GRADE method was used, evidence profiles were

created for the outcomes from the outpatient hysteroscopy reviews and compared to the

gradings in the original guideline.

This chapter examines how assessment of the quality of evidence differs between the GRADE

and SIGN guideline methods and discusses whether the GRADE method offers significant

advantages that might benefit the RCOG Green-top guidance.

Methods for comparing SIGN and GRADE for assessing quality of evidence

Using the seven systematic reviews of outpatient hysteroscopy (described in previous

chapters), a guideline was written entitled ‘Best Practice in Outpatient Hysteroscopy’ on

behalf of the British Society of Gynaecological Endoscopists (BSGE). The guideline was

developed with two NHS consultants and used the RCOG methods of guideline development

99

(119) which grades the quality of evidence and recommendations using the Scottish

Intercollegiate Guideline Network (SIGN) model (125) (Table 6.1). The final guideline was

published as a greentop guideline by the RCOG in association with the BSGE. The full

guideline can be reviewed in Appendix 4

Table 6.1 SIGN method for classification of evidence for clinical guidelines (120)

Classification of evidence levels

1++ High-quality meta-analyses, systematic reviews of randomised controlled

trials or randomised controlled trials with a very low risk of bias

1+ Well-conducted meta-analyses, systematic reviews of randomised

controlled trials or randomised controlled trials with a low risk of bias

1- Meta-analyses, systematic reviews of randomised controlled trials or

randomised controlled trials with a high risk of bias

2++ High-quality systematic reviews of case–control or cohort studies or high-

quality case–control or cohort studies with a very low risk of confounding,

bias or chance and a high probability that the relationship is causal

2+ Well-conducted case–control or cohort studies with a low risk of

confounding, bias or chance and a moderate probability that the

relationship is causal

2- Case–control or cohort studies with a high risk of confounding, bias or

chance and a significant risk that the relationship is not causal.

3 Non-analytical studies, e.g. case reports, case series

4 Expert opinion

Grades of Recommendation

A At least one meta-analysis, systematic reviews or RCT rated as 1++, and

directly applicable to the target population; or A systematic review of

RCTs or a body of evidence consisting principally of studies rated as 1+,

directly applicable to the target population and demonstrating overall

consistency of results.

B A body of evidence including studies rated as 2++ directly applicable to

the target population, and demonstrating overall consistency of results; or

extrapolated evidence from studies rated as 1++ or 1+

C A body of evidence including studies rated as 2+ directly applicable to the

target population, and demonstrating overall consistency of results; or

extrapolated evidence from studies rated as 2++

D Evidence level 3 or 4; or extrapolated evidence from studies rated as 2+

Recommended best practice based on the clinical experience of the

guideline development group

100

On reviewing Table 6.1 it is clear that by using the SIGN classification method, even the most

poorly conducted randomised controlled trial (RCT) can at worst be classified as 1- and any

recommendations that arise from this evidence will be graded as an A or B. Similarly, a

rigorous observational study can at best be judged as 2++ and recommendations arising from

it classified as C level evidence. This system puts the emphasis on trial design and not on trial

quality.

The GRADE guideline method examines eight determinants of quality; five which can lead to

downgrading of evidence ( limitations of the study design, inconsistency, indirectness,

imprecision and publication bias) and three which can lead to upgrading of the evidence

(strong association, evidence of a dose-response gradient and possible dilution due to

confounding). Randomised controlled trials start as high quality and are downgraded as

appropriate and observational studies start as low quality and are upgraded as appropriate.

The results of this evaluation are then tabulated as GRADE evidence profiles. The studies

within the meta-analysis or the individual study that provides the evidence for

recommendations must be evaluated for each of the outcomes and a decision made whether

there is enough concern to downgrade the quality of the evidence. The hysteroscopy guideline

only used randomised controlled trials and so in this evaluation the emphasis was on

downgrading.

’Limitations of the study design’ relates to the methodology of the studies and downgrades

the quality if there are concerns about randomisation method, allocation concealment,

blinding, reporting of pre-stated outcomes and loss to follow-up or inappropriate analysis (e.g.

not analysed as intention to treat). When ‘inconsistency’ is considered, studies are

101

downgraded if there is heterogeneity that cannot be explained or if the point estimates vary

widely with little overlap of the confidence intervals. When the systematic review or studies

assess a different population, use surrogate outcomes or comparisons are not head-to-head,

the evidence should be downgraded for ‘indirectness’. ’Imprecision’ considers the size of the

study population to assess whether it is adequate to answer the point in question and examines

the width of confidence intervals to assess if the clinical action would differ if the true point

estimate lies at the upper or lower limit of the confidence interval. The final criterion for

downgrading is ‘publication bias ‘which is likely if there are only a few small studies,

duplicate publications or the studies are unrepresentative (all in English, all find a significant

effect). Studies can be upgraded if there is a strong association (relative risk greater than 2 or

less than 0.5), if there is evidence of a dose-response gradient or if the presence of

confounding factors may have diluted the perceived result. However, it is recommended that

studies with limitations are not upgraded.

To examine how the quality of evidence in the hysteroscopy guideline would be graded if the

GRADE system was used evidence profiles were drawn for each of the recommendations in

the hysteroscopy guideline, that had been directly drawn from the systematic reviews

(chapters 2-5) using GradeProfiler (126). The GRADE tables are displayed in Appendix 5

with the recommendations and quality gradings from the original guideline. Each profile was

made up of a number of elements which addressed different aspects of the clinical question

e.g. when asking which type of distension media to use, five elements were investigated, (i)

pain, (ii) shoulder tip pain, (iii) vasovagal episodes, (iv) procedure time and (v) unsatisfactory

view. This GRADE table is displayed below as an example (Table 6.2)

102

Table 6.2 GRADE evidence table examining the evidence for using different types of distension medium for outpatient hysteroscopy

Quality assessment No of patients Effect

Quality Importance

No of studies

Design Risk of

bias Inconsistency Indirectness Imprecision

Other considerations

Normal saline

Carbon dioxide

Relative (95% CI)

Absolute

Pain score (Better indicated by lower values)

6 randomised trials

very serious

1

serious2 no serious

indirectness no serious imprecision

none 540 569 - SMD 0.34 higher (0.12 lower to 0.8 higher)

VERY LOW

CRITICAL

Shoulder tip pain

5 randomised trials

very serious

1

no serious inconsistency

no serious indirectness

no serious imprecision

none 9/472 (1.9%)

55/448 (12.3%)

OR 0.19 (0.09 to 0.4)

97 fewer per 1000 (from 70 fewer to 110 fewer)

LOW

CRITICAL

Vasovagal episodes

3 randomised trials

very serious

1




reporting bias3 5/210

(2.4%) 18/210 (8.6%)

OR 0.31 (0.12 to 0.82)

57 fewer per 1000 (from 14 fewer to 75 fewer)

VERY LOW

IMPORTANT

Procedure time (Better indicated by lower values)

4 randomised trials

very serious

1


4



none 434 401 - SMD 1.32 lower (1.48 to 1.17 lower)

LOW

IMPORTANT

Unsatisfactory hysteroscopic view

1 randomised trials

very serious

1



serious5 reporting bias

6 4/100

(4%) 19/100 (19%)

RR 4.75 (1.61 to 16.4)

712 more per 1000 (from 116 more to 1000 more)

VERY LOW

CRITICAL

1 Poor randomisation methods (quasi-randomised in some cases) and lack of allocation concealment in most studies. Neither patient nor operator was blinded but this would not be possible. 2 High heterogeneity that cannot be explained by differences in technique. 3 Outcome reported in just three of the six identified trials. 4 High heterogeneity but all studies in favour of normal saline and heterogeneity can be explained by the two studies that don't using a speculum showing the largest effect size. 5 Single study therefore imprecise. 6 This is the only study that reports a difference.

103

The importance of each aspect was determined by clinicians and not patients; however this

judgment does not contribute to the assessment of evidence quality and therefore was not felt

to be a significant omission. A score was then allocated to the quality assessment in its

GRADE and SIGN forms, with the highest quality rating awarded four points and the lowest

(see Table 6.3), one point. The difference between the two methods could then be quantified.

Table 6.3 Score allocated to the quality of evidence gradings as determined by SIGN and

GRADE

Score SIGN GRADE

4 A High

3 B Moderate

2 C Low

1 D Very low

To investigate which of the five aspects of the GRADE method (study limitations,

indirectness, inconsistency, imprecision and publication bias) were most often chosen as

reasons to downgrade the quality of the evidence, points were allocated to each aspect in the

GRADE evidence profiles as detailed in Table 6.4. When there was ‘no risk’ of bias the score

allocated was ‘0’, When there was ‘serious risk’ of bias a score of ‘1’ was allocated and if

there was ‘very serious’ risk of bias a score of ‘2’ was given. The aspect ‘publication bias’

can only be downgraded to ‘serious’ and so in this category a score of ‘2’ was not possible.

There were 37 elements examined using evidence profiles and so the maximum point

allocation for each aspect was 74 (37 x 2), except for ‘publication bias’, which was just 37.

104

Table 6.4 Scores applied to the downgrading aspects of the GRADE profile elements

GRADE aspect

Risk Score

Study limitations

No serious risk

Serious

Very serious

0

1

2

Inconsistency

As for ‘study limitations’

Indirectness


Imprecision


Publication bias

Undetected

Strongly suspected

0

1

Results of the comparison of SIGN and GRADE for assessing quality of

evidence

The differences between the allocated quality scores for the evidence behind each

recommendation were plotted graphically (Figure 6.1).

Figure 6.1. Quality rating according to SIGN and GRADE methodology and the

difference between the two scores.

105

Eleven of the recommendations from the SIGN guideline were converted into GRADE

evidence profile tables with 37 elements. Figure 6.1 above shows the points allocated to each

recommendation by SIGN and GRADE. It then shows the value when the GRADE rating is

subtracted from the SIGN rating to show the difference between the two methods. A positive

results shows that SIGN estimates the quality as higher than GRADE, a result of zero shows

that the assessment was the same in both systems and a negative result shows that GRADE

assessed the evidence quality as higher than SIGN. The figure illustrates that SIGN has a

tendency to grade evidence as higher quality than GRADE because all of the

recommendations were allocated a positive results when the scores were subtracted. None of

the recommendations were assessed as being the same level of quality by both of the

guidelines. The points difference ranged from one to three points.

Figure 6.2. Reasons for downgrading the quality of evidence for guideline

recommendations when using the GRADE system.

Figure 6.2 shows the contribution of each quality aspect to downgrading of the evidence. It

gives an overview of the quality of the evidence used throughout the guideline. It is clear that

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

% not selected

% selected

106

the most common reasons for downgrading evidence were ‘study limitations’,’ imprecision’

and ‘publication bias’. This is consistent with studies which have small populations and are

poorly conducted and also reflects the paucity of data available. However, Figure 6.2 also

suggests that the studies addressed the right questions and were reasonably homogenous as

‘indirectness’ and ‘inconsistency’ were rarely selected as reasons to downgrade the evidence.

Evidence for all recommendations was downgraded and so upgrading was not appropriate.

Discussion of the comparison of SIGN and GRADE for assessing quality of

evidence

Guidelines need to make clear recommendations. If the recommendations are open to

interpretation they become unfit for purpose and rather than guiding practice can cause

confusion and hesitancy. However, these recommendations can only be based upon the

available evidence and it is important that the quality of the evidence is taken into account

when making recommendations about health care. This evaluation of two different systems of

assessing available evidence to formulate guideline recommendations showed that the

GRADE and SIGN methods produced varying estimates of the quality of evidence. This may

not be surprising given that the systems adopt different methodologies. The GRADE

approach aims to demonstrate the available evidence in a transparent fashion that allows the

users of the guideline to understand why recommendations are afforded their weight. In

contrast the SIGN guideline system uses an alternative method of grading the quality of

recommendations which is based on the hierarchy of published data, with systematic reviews

and meta-analysis at the top, followed by randomised controlled trials and then observational

studies.

107

Guideline developers have focused upon the quality of available evidence to inform

recommendations but other considerations need to be taken into account especially as often

the evidence available is of poor quality. These factors include the potential for an

intervention to cause harm and its cost. For example, if we have an intervention that is cheap

and unlikely to be harmful, we may recommend its use (if there is no better alternative) even

if the quality of the studies showing benefit are of poor quality. However, if the intervention

had the potential to cause significant harm or was prohibitively expensive we would be

unlikely to recommend it. Thus, it is clear that when formulating recommendations from the

available evidence, the benefits and harms of the intervention need to be taken into account in

addition to the quality of the evidence.

The GRADE method considers these additional aspects when making recommendations but

they do not form part of the quality of evidence judgment. SIGN rates quality according to

study design with little consideration of the outcomes or publication bias. This is a simple

system in which the user identifies the trial design and decides if there is any risk of bias.

Although this method is subjective the classifications system dictates that all

recommendations from RCT’s or meta-analysis of RCT’s will receive a high quality grading

of A or B even if there is a high risk of bias. Conversely, GRADE focuses less on study type

and considers multiple aspects of the available data when assessing quality. It is designed to

be standardised and systematic (121;127;128) implying that if different people were presented

with the same body of evidence and asked to produce GRADE evidence tables they would all

come to the same conclusions regarding quality. However, a Spanish group examined the

introduction of the GRADE system into oncology guidelines and looked at how different

108

groups of clinicians rated the same evidence. For the same aspect of care, three groups rated

the body of evidence as ‘low’ quality, nine as ‘moderate’ quality and three as ‘high’ quality

(129). This casts doubt about the reproducibility of the GRADE system and is likely to be due

to factors inherent to the process.

The GRADE system is complicated (130) and time-consuming (128) because users must

consider multiple aspects during their quality assessment, each of which is open to some

degree of subjectivity. How does one decide whether the limitation identified is downgraded

to ‘serious’ or ‘very serious’? This is probably based on the individual’s prior knowledge and

experience of working with clinical studies. Users need to have some academic training to

understand the concepts behind each of the quality categories. Whilst most clinicians

understand different methods of randomisation and what is meant by ‘allocation

concealment’, when it comes to the more complicated criteria which require assessment of the

statistics or indicators of heterogeneity it is understandable that many would feel out of their

depth. Secondly, these judgments are also affected by the user’s expectations of trial design

and quality. Whilst an academic would prefer that the most rigorous methods are used at all

times, a clinician might recognises that the practicalities of running a study dictate that the

ideal methods cannot always be used if they are not pragmatic and would have jeopardised

trial recruitment. In these cases the clinician may feel that the best methods for the studies in

question were selected and that downgrading is not warranted, whilst the academic may

decide to downgrade to ‘very serious’ risks of bias. In the SIGN method this subjectivity will

only downgrade the quality of the recommendations to a B at worst but in the GRADE

system, if this subjectivity is present in multiple aspects of the quality assessment the data

could easily be assessed as high / moderate quality by one assessor and low / very by another.

109

Thus a criticism that could be justifiably levelled at the GRADE system is that it appears to

have been developed by academics for academics. If the increased complexity of GRADE

compared to other methods results in clinical guideline developers failing to understand the

process clearly (131) then this will limit GRADE from achieving the objective of being

standardised. One potential solution is to produce alternative methods of data display to make

guidelines more user friendly for clinicians, who are ultimately the end users. Clinicians may

find the guidance less confusing and easier to assimilate if only the outcome of the quality

assessment was presented to them but they understood that this judgment was made after a

rigorous assessment, using the GRADE methodology. Thus, the intimidating GRADE

evidence profile tables, displaying cumbersome volumes of data (132) along with the

rationale for upgrading or downgrading evidence quality could be dispensed with.. This

would be reminiscent of the SIGN guidelines which display the result of quality assessment

but don’t explain the reasoning for reducing (-) or increasing (++) the quality within the text.

However, as already discussed, the SIGN classification, which uses numbers and letters, does

not reliably convey the quality of evidence. Graphical representation of the GRADE evidence

profiles may prove more useful than the tables as it allows clinicians to quickly interpret the

data and make a judgment at a glance, without having to read lots of text (133). In addition

graphical display enables quick and easy comparison of the data quality for competing

strategies. Graphical display of GRADE data using radar charts and traffic light systems have

been suggested (132;134) although they are yet to be validated.

The current study is a comparison of two methods for assessing evidence quality for just one

guideline. Future research should assess other guidelines in a similar way to evaluate whether

110

GRADE is indeed associated with more conservative grading of evidence quality and whether

the magnitude of differences are similar to the current study . If this is the case, then research

addressing the clinical significance of using the GRADE system upon existing

recommendations would be needed. In addition further research should examine how

clinicians interpret quality gradings when they are displayed in different formats, including

tables, diagrams and codes.

In summary, when comparing the GRADE and SIGN methods for grading the quality of the

evidence used in the guideline for ‘Best Practice in Outpatient Hysteroscopy’ the two

different systems produced varying estimates of the quality of evidence. Furthermore,

GRADE appeared to rate the evidence less highly than SIGN. It is evident from this

assessment that neither GRADE nor SIGN offer a perfect solution to the assessment of

evidence quality when it comes to writing clinical guidance. Whilst SIGN may be too

simplistic and under or overestimate quality, GRADE is too complex which results in it being

used incorrectly and the judgments not being reproducible. The GRADE system in its current

form may need to be adapted so that results are displayed in a more user-friendly style,

possibly by the incorporation of diagrams to present the evidence quality. A hybrid system

which uses GRADE to assess the quality but displays only the result of the quality assessment

in the guideline may prove to be the most manageable form of guidance for clinicians.

Furthermore, high quality evidence is far from ubiquitous, and so the benefits, harms and

costs of interventions should also have an influence upon the degree of evidence quality

required in order to make a particular recommendation.

111

CHAPTER 7

INTRODUCTION TO THE ECONOMIC ANALYSIS OF

DIAGNOSTIC STRATEGIES FOR THE INVESTIGATION OF

HEAVY MENSTRUAL BLEEDING

The following economic analysis forms a small part of a report that was written for the

National Institute for Health Research, Health Technology Assessment programme. The

report investigated the cost-effectiveness of diagnostic tests in the management of all types of

abnormal uterine bleeding. It is due to be published in late 2013.

Heavy Menstrual Bleeding

HMB affects 1 in 5 women of reproductive age, with 5% of women aged 30-49 consulting

their General Practitioner each year because of the condition and accounting for a third of all

gynaecology referrals (135). The overall prevalence of HMB in England and Wales has been

estimated at 1.5 million women (136). The number and cost of consultations and treatments

impose substantial demands on health service resources (137;138). Moreover, HMB can cause

significant distress to women by affecting their performance at work as well as their social

activities, and imposes a substantial adverse impact upon health-related quality of life

(HRQL) (139-141). The National Institute for Clinical Excellence in their 2007 guideline into

the management of heavy menstrual bleeding (HMB) recommended the condition be defined

as “excessive menstrual blood loss which interferes with the woman’s physical, emotional,

112

social and material quality of life, and which can occur alone or in combination with other

symptoms” (142). This clinical definition is the most useful one as using objective

measurement, with loss of > 80 ml of blood per cycle considered definitive of HMB (143), is

impractical. More applicable semi-objective measurement, using pictorial blood loss

assessment of sanitary ware (144;145) as a surrogate for objective measurement, has been

tried but the correlation between objective and semi-objective quantification has been

questioned (144). In any case, objective quantification of menstrual loss does not correlate in

many cases with a woman’s subjective complaint of HMB (146-148).

Causes of heavy menstrual bleeding

Heavy menstrual bleeding has been reported to be caused by a variety of underlying

pathologies (142). However, whilst many conditions have been linked to HMB, in practice

most cases are attributed to fibroids, endometrial pathology or dysfunctional uterine bleeding

(DUB) and subsequent treatment is dictated by the presence or absence of these conditions

(Table 7.1).

113

Table 7.1 Causes of heavy menstrual bleeding

Cause

Definition

Dysfunctional uterine bleeding The occurrence of irregular or excessive uterine bleeding in the

absence of identifiable organic pathology (142).

Uterine fibroids Smooth-muscle tumours of the uterus, generally benign

(>99%). They vary in size from millimetres to tens of

centimetres, and are associated with heavy periods, pressure

symptoms and occasionally pain. They are responsive to

oestrogen and progesterone, so tend to shrink after the

menopause (142).

Endometrial pathology:

Polyps

Hyperplasia

Cancer

Focal outgrowths within the uterine cavity, containing a

variable amount of glands, stroma and blood vessels which

influence their macroscopic appearance. Usually benign with

less than 1% frankly malignant (6).

Endometrial hyperplasia is a proliferation of endometrial glands

with structural abnormalities and crowding. If the nuclei exhibit

cytological atypia the hyperplasia is classified as atypical (149)

and is considered pre-malignant.

Well-differentiated carcinoma is distinguished from atypical

hyperplasia by the presence of endometrial stromal invasion

(149). Both conditions are rare in premenopausal women.

Diagnosis of heavy menstrual bleeding

The current National Institute for Health and Clinical Excellence (NICE) guideline advocates

full gynaecological examination and taking a full blood count to exclude anaemia (142). This

guideline (142) recognises the need for diagnostic tests to evaluate the uterus, namely

endometrial biopsy, ultrasound scan and hysteroscopy in specific cases. These tests are

described in Table 7.2.

114

Table 7.2 Description of currently used tests for the diagnosis of uterine pathology

Test Description (adapted from CG44 NICE

Heavy Menstrual Bleeding Guideline

2007 (142))

Capability

Tests primarily for detecting structural abnormalities

Transvaginal

Ultrasound

(TVS)

A method of imaging the genital tract in

women using high-frequency sound waves

that bounce off body structures to create a

picture on a screen. The ultrasound

transducer is inserted directly into the

vagina so that it closer to pelvic structures

than with the conventional transabdominal

technique.

Diagnoses

endometrial, focal

intracavity (polyps,

submucous

fibroids)

myometrial and

adnexal pathology

Saline infusion

ultrasound

This is an extension of transvaginal

ultrasound. Prior to the ultrasound, saline is

injected into the uterine cavity to distend it

and provide better visualisation of the

anatomic structures within it by providing a

liquid-solid interface.

As for TVS but

with enhanced

diagnosis of focal

intracavity

pathology

Outpatient

hysteroscopy

A hysteroscopy is an examination of the

inside of the uterus using an endoscope. The

hysteroscope is carefully passed through the

vagina and cervix and into the uterine cavity

to allow direct visualisation. During the

procedure a biopsy may be taken for

examination.

Diagnoses

endometrial and

focal intracavity

pathology

Tests primarily for detecting histological abnormalities

Endometrial

biopsy

A sample of endometrial tissue is obtained

blindly using a sampler which is passed

through the cervix and uses suction to obtain

the tissue. This can then be histologically

examined

Diagnoses

endometrial

pathology at a

histological level.

Dilatation and

curettage

A procedure performed under general

anaesthetic in which the cervix is

mechanically dilated and a curette is

introduced into the uterine cavity to scrape

away a sample of endometrial tissue.

As for endometrial

biopsy

115

Current diagnostic pathways for heavy menstrual bleeding

It is only in the last 25 or so years, that evaluation of the uterine cavity in women with HMB

symptoms has moved on from dilatation of the cervix and curettage of the endometrium lining

from the uterine cavity (D&C). This test is now used only in exceptional circumstances as it

requires general anaesthesia and has been superseded by outpatient endometrial biopsy, which

obtains endometrial tissue samples for histological analysis in a convenient outpatient setting

without the need for anaesthesia (150;151). Moreover, the development of high resolution

transvaginal ultrasound (TVS) has allowed female pelvic structures, including the uterus, to

be visualised. The ‘inside’ of the uterus i.e. the uterine cavity, cannot normally be seen

without effecting distension using a fluid or gaseous medium to separate the opposing walls

of the uterus. This potential limitation of TVS has been overcome by the advent of saline

infusion sonography (SIS) (152-154) and outpatient hysteroscopy. These tests, described in

Table 7.2, provide different, albeit overlapping information and diagnostic accuracy varies

according to the particular pathology under scrutiny. NICE guidance from 2007 recognised

that “…particular investigative methods were better for identifying certain types of pathology

than others.”

Thus, the availability of different, easy to use, miniature and increasingly portable ‘bed side’

tests has created uncertainty as to how best to employ them. This is particularly true in HMB

where different aetiologies need to be considered and the preceding clinical history, and more

often than not the examination too, are unable to predict causation with accuracy. As a result

current testing is eclectic, depending upon the preferences of individual clinicians and the

availability of services locally.

116

Literature review of cost-effectiveness studies for the diagnostic work up of

heavy menstrual bleeding

A systematic search was performed of Medline and Embase electronic bibliographic

databases using the terms ‘heavy menstrual bleeding’ and ‘cost-effectiveness’ along with their

MeSH terms. 350 articles were identified once duplicates had been removed. Three relevant

economic evaluations of diagnostic tests used for evaluating HMB were identified. One

evaluation took place alongside an RCT (155) and the other two were economic modelling

studies (142;156).

Cost-effectiveness was examined in a RCT conducted between 1999-2001 in Scotland

comparing three outpatient diagnostic tests (outpatient biopsy, ultrasound and hysteroscopy)

for the evaluation of abnormal uterine bleeding (AUB) in certain test combinations (155).

Women were split into high, moderate and low risk groups for endometrial cancer. Resource

use tended to be higher in the moderate- and low-risk women, because of the need to manage

their persistent abnormal bleeding symptoms. Minimal difference in cost-effectiveness was

found between investigation options in the high-risk group (postmenopausal), with the option

involving hysteroscopy being marginally better than ultrasound (£88/woman, compared with

the other options). The most cost-effective investigation in the moderate-risk group was

biopsy alone (saving £128–212/ woman better) and in the low-risk group ultrasound (£74–

452/woman better).

The mixed population of women with AUB, i.e. women of reproductive age with HMB and

postmenopausal women with unexpected vaginal bleeding, limits clinical inferences because

117

the aim of investigation of women with PMB is to exclude endometrial cancer whereas in pre-

menopausal women it is to optimise management of benign pathology. The authors of this

RCT (155) highlight this themselves by stating “…in future research into the evaluation and

management of AUB, postmenopausal women should be studied separately from

premenopausal women with menstrual bleeding problems”. Furthermore, the primary end-

point defining ‘effectiveness’ was based upon the premise that a satisfactory diagnosis must

have been reached once no further investigation had been carried out, as identified by

retrospective case note review. Clearly such an indirect assumption of effective diagnosis,

whilst expedient, is unlikely to be a reliable or valid measure of effectiveness and does not

take account of patient-centred outcomes (e.g. satisfaction, reduction in bleeding, survival

etc.).

As well as economic data from effectiveness studies, an alternative approach to assessment of

cost-effectiveness of diagnostic testing is to employ decision-analytic modelling. Two

economic evaluations of diagnostic testing in HMB using decision analytic modelling have

been published (142;156). The first of these analyses was conducted from the perspective of

the Dutch Health Care system and compared the percentage of patients treated successfully

and the cost of six strategies for the evaluation of HMB: (0) hormonal treatment, (I) treatment

of all patients with balloon ablation, (II) TVS and therapeutic hysteroscopy, (III) TVS, SIS

and therapeutic hysteroscopy, (IV) SIS and therapeutic hysteroscopy, and (V) diagnostic

hysteroscopy and therapeutic hysteroscopy. Hormonal treatment was considered to be the

reference strategy to which the five strategies were compared. The study found that the

strategy starting with SIS (IV) and the strategy with diagnostic hysteroscopy (V) revealed the

highest number of patients treated successfully for HMB. However, the diagnostic strategy

118

based upon initial evaluation with SIS was the most cost-effective strategy for successful

treatment of HMB, especially when the prevalence of intracavity pathology (polyps,

submucous fibroids) was high. Study weaknesses limit to some degree the validity and

stability of these findings. These included problems with construction of the decision model

(limited pathologies were taken into account e.g. diagnosis of intramural fibroids and

endometrial disease were not considered), the use of outmoded treatments (e.g. use of long

term systemic progestogen) and overlooked ambulatory outpatient based treatment. Failure

rates of testing were unaccounted for, the precision and quality of data sources used for

estimating test accuracy were questionable and the definition of therapeutic effectiveness was

unclear. The findings of the analysis were sensitive to changes in the key assumptions and so

the robustness of the results is questionable.

The other decision-analytic model was developed to examine the cost- effectiveness of three

imaging techniques; TVS, SIS and hysteroscopy from an NHS perspective (142). The model

showed that TVS was more accurate and less costly than either SIS or hysteroscopy. For a

cohort of 1000 women examined for the presence of structural abnormalities, ultrasound

generated 810 correct diagnoses at a cost of £107,490 compared with 735 correct diagnoses at

a cost of £145,110 using saline infusion sonography and 696 correct diagnoses at a cost of

£209,720 using hysteroscopy. Although the economic analysis was conducted from an NHS

perspective, the general applicability of the model is limited due to its simplistic construction.

Women were assumed to have one of two health states: ‘no intrauterine pathology’, or ‘any

intrauterine pathology’ and the outcome measure chosen was ‘correct diagnosis’. This was a

pragmatic choice given the scope of the guideline (142) such that it was not possible to

construct a model designed to take into account the range of pathologies under consideration

119

for HMB, and the associated range of treatment pathways. The impact on cost-effectiveness

of women falsely diagnosed was not considered (the model did not follow women beyond an

initial diagnosis) so that the model does not reflect the true longer term costs and outcomes

associated with each diagnostic method. Moreover, diagnosis was restricted to one test,

whereas this does not reflect contemporary practice where multiple testing is likely, either

conducted simultaneously or conditional on previous test results.

The relative dearth of comprehensive diagnostic cost-effectiveness data in women with HMB

reflects the complexity of care pathways (i.e. the varied outpatient tests available, the range of

uterine pathologies) and the relatively recent introduction of minimally invasive, ‘ambulatory’

or ‘outpatient’ treatments.

Current treatment of heavy menstrual bleeding

Medical therapy

According to the recent NICE guideline on HMB, medical treatments should be considered

where (i) structural and histological abnormalities of the uterus have been excluded; (ii) for

fibroids < 3 cm in diameter which do not appear to distort the cavity of the uterus or (iii)

where future fertility is required (142). The first line recommended medical treatment by

NICE is the levonorgestrel-releasing intrauterine system (LNG-IUS or Mirena®, Bayer

Healthcare Pharmaceuticals, Pittsburg, PA, USA) which is an effective non-surgical treatment

for HMB, is reversible, contraceptive and fertility sparing. In the majority of cases, the device

is fitted easily within a few minutes in the outpatient setting. Endometrial proliferation is

suppressed as a result of local release of the synthetic progestogen levonorgestrel (LNG) and

120

this leads to a reduction in estimated menstrual blood loss of up to 96% by 12 months, with

up to 44% of users reporting amenorrhoea, (157;158) at a cost which is a third of that for

hysterectomy (159). However, the LNG-IUS can lead to troublesome break through bleeding

and vaginal discharge in some women, causing early discontinuation of the device. The LNG-

IUS works effectively in a relatively normal sized uterus (<11cm sound length) without

distortion by focal pathology (160) i.e. in DUB or the presence of small uterine fibroids (<12

week uterine size) which do not encroach into the uterine cavity (161). Local release of

progestogen can also reverse endometrial hyperplasia without atypia (162;163). Thus the

LNG-IUS is applicable for most aetiologies of HMB with the exception of focal pathologies

distorting the uterine cavity, large uterine fibroids (>12 weeks size) or in the presence of

atypical hyperplasia or endometrial cancer.

Surgical treatment

Long-term medical treatment with the LNG-IUS is unsuccessful or unacceptable in many

cases and surgical alternatives may be required (164). Traditional surgical treatment of HMB

refractory to medical intervention has been with hysterectomy, but now removal of the uterus

is generally restricted to women where conservative, uterine sparing surgical procedures have

been unsuccessful or in the presence of large fibroids, atypical hyperplasia and endometrial

cancer. Endometrial ablation is a technique where a semi-automated device is placed in the

uterine cavity and thermal energy is applied to the endometrium and superficial myometrium

to achieve a uniform, global and permanent destruction of the endometrial lining, thereby

inducing amenorrhea or substantially reducing menstrual blood loss (6). Various modalities

are available including fluid filled thermal balloons, free circulating warmed saline, bipolar

radiofrequency ablative systems and cryotherapy. Prior to the automated techniques,

121

endometrial ablation was performed with a resectoscope, however this technique is now used

in rare circumstances, having been replaced by the safer and cheaper (165) second generation

methods. Endometrial ablation is recommended as a second line treatment when fertility is

not desired and medical treatment has failed in DUB but can also be used in the presence of

uterine fibroids when there is a relatively normal sized and shaped uterine cavity.

Hysteroscopic resection of focal intracavity lesions including polyps (166;167) and

submucous fibroids (111;168-171) has been shown to improve HMB symptoms found in

association with these pathologies. Hysteroscopic removal is standard practice in the UK

(172) and in the case of polyps can usually be achieved in the outpatient setting (173-176).

The procedures involve the use of electrosurgical cutting electrodes placed down a small

operative working channel in the hysteroscope or a formal hysteroscopic resectoscope using a

larger loop electrode.

In the presence of significant fibroids associated with an estimated uterine size of >12 weeks,

and when retention of fertility is not required, hysterectomy is usually recommended. Uterine

artery embolisation (UAE), is a less invasive, uterine sparing, interventional radiological

intervention (177). This procedure is normally restricted to women with medical or surgical

risk factors for open surgery. Myomectomy (removal of fibroids with conservation of the

uterus) is sometimes offered but as it is as invasive as hysterectomy but less effective (178),

the technique is generally reserved for women wanting to retain their fertility or to improve

fecundity in those women with subfertility associated with a large fibroid uterus.

122

Defining treatment success in heavy menstrual bleeding

Menstruation is normal for women of reproductive age and so defining ‘successful treatment’

can be problematic. The primary aim of treating HMB is not to eradicate bleeding altogether,

although some interventions do induce amenorrhoea, but to ameliorate bleeding symptoms to

a tolerable level. As already discussed, objective measurement of reduction in menstrual

bleeding is impractical and lacks relevance. Many studies have tried to measure the impact of

interventions upon patient’s quality of life and / or satisfaction with treatment outcome.

Health related quality of life

Generic health related quality of life measures have been used, but many have not been

validated for use in HMB and fail to capture the cyclical nature of the symptom (179;180). In

addition, they lack sensitivity as most women suffering with HMB are otherwise healthy and

can continue to function in most generic health domains during menstruation (140;141).

Condition-specific measures have been developed for HMB but either only assess surgical

interventions (181) (as opposed to medical ones) or have been infrequently used which has

limited a full assessment of their inherent psychometric qualities (182).

Satisfaction

Patient satisfaction is widely used as a primary outcome measure in studies of treatments for

HMB and guidelines (136;142;183;184). Satisfaction is a subjective concept and represents

the extent to which a service meets the users’ expectations. A variety of questions and scales

have been used to elicit satisfaction with treatment in HMB studies but this lack of uniformity

has precluded meta-analysis of data across studies (136;166). Furthermore, the validity of

123

current patient satisfaction measures is questionable in light of the lack of published studies

examining their development and application in HMB.

124

CHAPTER 8

METHODS FOR THE ECONOMIC ANALYSIS OF



Construction of the decision model

A clinically informed cost effectiveness model was drawn as a decision tree using Treeage

software (185) to reflect current service provision for the diagnostic work up of women

presenting with heavy menstrual bleeding. The tree was constructed to examine the

effectiveness of different diagnostic testing strategies for women referred to secondary care

by their general practitioner. The tests evaluated were transvaginal ultrasound (TVS), saline

infusion sonography (SIS), global endometrial biopsy (EBx) and outpatient hysteroscopy

(OPH). The tree structure was informed by clinical input. As there is no consensus regarding

how best to investigate women with HMB, initial investigation utilising all tests either alone

or in combination were included in the model. Therefore the tree consisted of the four tests

available deployed in isolation or in various clinically relevant combinations following initial

presentation. The need for any additional subsequent tests was conditional upon the preceding

test result(s). This resulted in the formation of eleven clinically relevant, alternative testing

strategies. In addition, two scenarios were developed where testing was dispensed with and

treatment of HMB instituted immediately regardless of diagnosis. The treatments chosen were

the most effective medical treatment (the levonorgestrel intrauterine system (LNG-IUS)) and

125

surgical treatment (hysterectomy). This allowed comparison of the various approaches to

diagnostic work up with the option of ‘no investigation’. In view of the fact that NICE

guidance (142) recommends the use of the LNG-IUS as first line treatment in HMB, this arm

was used as the base-case scenario to compare all other strategies against. An incremental

approach was used for reporting the results. Thus, in total there were 13 different scenarios

evaluated in the decision model (eleven testing and two treatment alone strategies) which are

listed below:

1. LNG-IUS alone

2. Hysterectomy alone

3. Outpatient hysteroscopy

4. Transvaginal scan

5. Endometrial biopsy

6. Saline infusion sonography

7. Outpatient hysteroscopy and endometrial biopsy

8. Transvaginal scan and endometrial biopsy

9. Saline infusion sonography and endometrial biopsy

10. Outpatient hysteroscopy and saline infusion sonography

11. Outpatient hysteroscopy and transvaginal scan

12. Saline infusion sonography, outpatient hysteroscopy and endometrial biopsy

13. Transvaginal scan, outpatient hysteroscopy and endometrial biopsy

126

Structure of the model

A series of decision trees evaluating various testing strategies for HMB were developed to

represent alternative decision options and their possible consequences. The trees explicitly

illustrate the patient pathway from suspected pathology underlying the clinical presentation

through to the outcome of testing, distinguishing between correct and incorrect diagnosis.

Then, conditional on the accuracy of the diagnostic testing strategy, the outcome of treatment

for HMB was analysed at one year post initial presentation. Disease prevalence, diagnostic

test accuracy, treatment effectiveness along with associated costs was used to populate the

relevant branches of the decision tree. The basic tree structure is illustrated in (Figure 8.1).

The thirteen trees representing the diagnostic testing options for HMB are detailed in

Appendix 6, however the trees themselves are too large to display completely so a branch of

one tree has been expanded as an example and a table has been included which details the

data from the remaining branches of the tree.

Deterministic results were obtained using point estimates of the parameters to estimate the

expected cost, outcome (satisfaction) and incremental cost-effectiveness (additional cost per

extra patient satisfied). The stability of the results was then tested through sensitivity analysis.

127

Figure 8.1 Example decision tree for evaluating the cost-effectiveness of diagnostic

testing in heavy menstrual bleeding

Clinical assumptions

The following section details the main clinical assumptions required to develop the economic

model. Senior gynaecologists (see Acknowledgments) who were selected based upon their

reputation and experience, were surveyed to ensure that the model structure and clinical

inputs reflected contemporary practice. The clinicians were contacted by email, as well as

telephone and face-to face interview. Initial correspondence concerned the structure of the

model and clinical management. The size of the model precluded it from being presented as a

whole, so only scenarios which appeared contentious were presented to the clinicians, asking

how they would manage the patients. When starting to populate the tree with data it became

clear that not all values could be identified from the literature. This was particularly true for

‘satisfaction’ values after inappropriate treatments (for example, women with large fibroid

128

uteri being treated with the levonorgestrel intrauterine system or endometrial ablation)

because these scenarios only arise when women are unknowingly treated with unsuitable

medications or procedures, hence the numbers are small and the data are not collected. In the

absence of a consensus view, the opinion of the majority was adopted or when data were

concerned the median values were used. Inevitably not all clinicians responded to all

questions.

Demographics, pathologies and treatments

It was assumed that women presenting with HMB had a mean age of 45 years and no wish for

future fertility. Forty-five years was selected for two reasons. Firstly, NICE recommends that

endometrial biopsy should not be performed routinely for women with HMB before this age

(142) as prevalence of endometrial premalignant or malignant disease is low. Secondly

because HMB is most prevalent in parous women over 40 years, most of whom have

completed their families (186) and are then eligible for all potential treatment options (the

desire for current or future fertility restricts treatment options in HMB i.e. avoidance of

hormonal contraceptive medical treatments or the surgical interventions endometrial ablation

or hysterectomy).

All women were assumed to have been referred from primary care and to have not previously

been seen for testing or treatment in secondary care. A single underlying aetiology was

assumed to be causative and concurrent pathologies were not considered. This assumption is

in keeping with the majority of HMB cases (142) and prevented unnecessary model

complexity.

129

The premise was that optimal treatment of HMB is dependent upon correctly diagnosing the

underlying cause so that appropriate, tailored treatment is arranged. The model was

constructed based upon the true underlying diagnosis. The true diagnosis was assumed to fall

within one of the following categories;

Intrauterine resectable pathology (endometrial polyps, submucous fibroids)

Fibroid uterus <12 weeks size (intramural and subserosal fibroids only)

Fibroid uterus >12 weeks size (intramural and subserosal fibroids only)

Endometrial disease (complex endometrial hyperplasia or endometrial hyperplasia

with atypia or cancer)

Dysfunctional uterine bleeding

When choosing the pathology categories, it was decided that endometrial polyps and

submucous fibroids should be grouped together as they are treated in the same way i.e.

hysteroscopic resection of focal lesions. The majority of polyps were assumed to be treated in

the outpatient setting (174;187;188) whereas the majority of submucous (intracavity) fibroids

were assumed to be treated under general anaesthesia (169;170), 70% of which also required

pharmaceutical endometrial down regulation with gonadatrophin releasing analogues (GnRH-

a) for three months prior to surgery following outpatient diagnosis. Only 70% were pre-

treated with GnRH-a because not all women would require or tolerate pre-treatment and not

all gynaecologists use it. A minority of women would undergo hysteroscopy and D&C under

general anaesthesia because the planned outpatient testing was unsuccessful. In these

circumstances, where a polyp or submucous fibroid was detected at hysteroscopy and D&C, it

was assumed that the focal lesion would be treated simultaneously; in the case of a

130

submucous fibroid this meant that GnRH-a pre-preparation of the endometrium would not

have been used.

Intramural (confined to the myometrium) and subserosal (extending beyond the myometrium

to distort the serosal surface of the uterus) fibroids were grouped according to size. This was

because smaller fibroids, which do not substantially distort the shape of the uterine cavity or

increase the uterine size beyond the equivalent size of a 12 week gravid uterus (the size at

which the uterus becomes palpable abdominally), do not contraindicate the use of LNG-IUS

or endometrial ablation (EA) (189-191); treatments which are successful in the majority of

women (184;192-194). Thus the presence of small fibroids without cavity encroachment is, in

practice, treated the same way as those women diagnosed with DUB (i.e. no identified

structural uterine pathology).

In contrast, large fibroids increasing the uterine size beyond 12 weeks size, tend to be treated

with invasive surgical interventions (abdominal hysterectomy or laparotomic myomectomy)

as the LNG-IUS and minimally invasive surgery (EA or hysteroscopic resection of focal

lesions) are either contra-indicated due to cavity size or ineffective (189;190). An alternative,

less invasive radiological intervention for large uterine fibroids is uterine artery embolisation

(UAE) but a randomised controlled trial of UAE and hysterectomy (195) has found no

statistically significant differences between them in terms of satisfaction and effectiveness.

Furthermore, hysterectomy is the gold standard definitive treatment and is more widely

available than UAE therefore it was chosen as the treatment for fibroid uteri beyond 12 weeks

size.

131

The majority of women with HMB have a benign, functional endometrium. However, over

proliferation of the endometrium can lead to endometrial hyperplasia, which in the minority

of cases (<5%) if left untreated, can result in the development of endometrial cancer (149).

Endometrial cancer is rare in pre-menopausal women, but hyperplasia is not infrequently

encountered as the result of anovulation and a relative excess of unopposed oestrogen

stimulating the endometrium. Histological assessment of the endometrium is the only way to

reliably diagnose endometrial hyperplasia and cancer and so EBx is mandatory where

suspected. Endometrial hyperplasia is treated hormonally with progestogens either delivered

systemically or more often now locally by fitting a LNG-IUS. Hysterectomy is recommended

where the hyperplastic process does not respond to progestogen treatment or in the presence

of cytological atypia. This is because the likelihood of developing malignant disease is

increased to around 25% in the presence of atypia (149). Endometrial cancer is generally

treated with hysterectomy with or without radiotherapy depending upon the stage and type of

cancer. Over 70% of endometrial cancers are diagnosed early and are confined to the uterus

so can be cured by timely hysterectomy (196). Given the rarity of malignant endometrial

disease in pre-menopausal women it was assumed that when endometrial cancer was

encountered it would be a well differentiated FIGO stage 1a endometrioid cancer, treated by

hysterectomy alone.

Dysfunctional uterine bleeding, although not a distinct pathology, is a diagnosis of exclusion

and the recommended first line medical treatment is the LNG-IUS (142). This was therefore

the chosen treatment in the model for DUB.

132

Setting and decision making

The clinical setting was assumed to be an efficient, contemporary ‘one-stop or ‘see and treat’

service run by a consultant gynaecologist. This setting meant that the expertise and

infrastructure were available to perform all stipulated tests at the same visit and that

therapeutic management could be implemented without unnecessary delay. When outpatient

treatment was indicated, such as fitting of a LNG-IUS or hysteroscopic removal of a uterine

polyp, then this was done at that visit. Interventions that required a general anaesthetic in a

formal theatre setting (e.g. hysterectomy, endometrial ablations) were scheduled for a later

date.

The results of all imaging tests would be available in real time to the senior clinician

performing the test. However, in the case of EBx, the result would be delayed whilst the

tissue sample was prepared, analysed and reported by the pathologist. Therefore when an EBx

was performed because endometrial disease was suspected, any treatment or treatment plan

could not be instigated immediately. However, when a testing strategy involved the initial use

of endometrial biopsy in combination with OPH which showed a probable benign cause for

HMB (normal appearance i.e. DUB or a focal lesions seen i.e. polyp or submucous fibroid),

treatment would be initiated at that first appointment. If endometrial disease was then

unexpectedly diagnosed once the biopsy result became available, an alternative treatment

would be instigated at a further appointment if it was felt to be more a more appropriate

treatment option.

Imaging tests (OPH, TVS and SIS) (197-199) can discriminate to some degree between

normal and abnormal endometria, but are unable to accurately differentiate between

histological subtypes of abnormal endometria; complex hyperplasia, complex hyperplasia

133

with cytological atypia or cancer (198;200). In keeping with clinical practice (opinion of

expert panel), when imaging tests diagnose an abnormal endometrium it was felt by the panel

of gynaecologists that no clinician would treat these suspected endometrial conditions without

a histological tissue diagnosis. Therefore, a confirmatory histopathological test was performed

if abnormal endometria were suspected by imaging; EBx will provide a result in 91% (151) of

women but the remaining 9% would need to undergo formal D&C under a general anaesthetic

as a day case because of failed procedures, non-diagnostic samples or patient preference.

Therefore the cost for a confirmatory test was a composite value calculated as 91% of the cost

of EBx plus 9% of the cost of D&C.

Formal hysteroscopy D&C under general anaesthesia, was considered a second line

diagnostic test and was restricted, for consistency, to the minority of women where initial

diagnostic testing was unsuccessful i.e. failure to complete the test.

Combination testing strategies and discordant results

When combinations of tests were used, the overall testing strategy was considered successful

only if all tests were completed successfully. Failure of one or both tests was considered a

failure of the testing strategy. This assumption seemed reasonable on clinical grounds and

from a modelling point of view because success of one test in a dual testing strategy would

simply replicate the analysis for the respective single test strategy in the model, rendering it

redundant.

When a testing strategy involved more than one test applied simultaneously, the decision trees

for each test were combined (appearing in series within the trees) to provide the additional

information associated with combined testing. The final diagnoses were based upon the

134

results from the combination of tests. Tests in agreement presented no uncertainty, but where

test results were modelled as being discordant, the panel of gynaecologists were asked to

recommend the favoured diagnosis.

False diagnoses

Table 8.1 lists the false negative diagnoses which were considered plausible for true

pathologies according to testing modality. The rationale underpinning the assumptions is also

described.

Treatment failure

Following diagnosis, the most suitable treatment was arranged (Table 8.2). Only one

treatment was considered for each diagnosis. As HMB is cyclical and patients would expect a

delay in symptom relief with the LNG-IUS and EA, outcomes could not be assessed for at

least six months. It was assumed that dissatisfied women would attend their general

practitioner (GP) and be referred back to secondary care to be reviewed by a gynaecologist,

who would undertake a further specific, second-line treatment (Table 8.2). The exception to

this strategy was dissatisfaction after initial treatment with hysterectomy because no further

treatment is possible in the absence of a uterus. These women were assumed to attend their

GP for a consultation only. Women that remained dissatisfied following a second treatment

were assumed to receive ‘rescue treatment’ consisting of a GP visit, a further hospital

gynaecology outpatient appointment and a total abdominal hysterectomy (unless

hysterectomy had been performed already in which case they were assumed to attend their GP

for a consultation only). Patients were assumed to undergo the first two treatments within a 12

month period.

135

Table 8.1 False diagnoses according to testing modality

True pathology

False diagnoses Rationale (Clinical consensus)

Transvaginal ultrasound

Intrauterine polyp or submucous fibroid Fibroids < 12 weeks

Endometrial Disease

DUB

Focal pathology can be easily missed by 2D imaging

without cavity distension. Endometrial polyps can appear

cystic and thus be mistaken for endometrial disease. A

small fibroid encroaching into the endometrial cavity

(submucous fibroid) could be erroneously considered

intramural forming part of a small fibroid uterus.

Fibroid uterus <12 weeks size Polyp / SMF

Endometrial Disease

DUB

Intramural fibroids may be wrongly diagnosed as

submucosal. Small fibroids could be missed and a

thickened, functional endometrium could appear

hyperplastic.

Fibroid uterus >12 weeks size Polyp / SMF

Fibroids < 12 weeks

Large fibroids would be rarely overlooked entirely, but it

is possible to underestimate their size or incorrectly

classify fibroid location

Endometrial disease Polyp / SMF

Fibroids < 12 weeks

DUB

A thickened hyperplastic or cancerous endometrium

could be misdiagnosed as containing a polyp. Small

fibroids may be incorrectly identified within the

myometrium. The endometrium may appear to be

normal.

Dysfunctional uterine bleeding Polyp / SMF

Fibroids < 12 weeks

Endometrial disease

A normal, thickened endometrium could be considered

falsely to be some form of endometrial disease

(hyperplasia or cancer) or focal lesion (e.g. folds of

normal endometrium mistaken for a polyp). Small

fibroids may be incorrectly identified within the

myometrium.

136

Table 8.1 continued

True pathology


Saline infusion sonography

Intrauterine polyp or submucous fibroid Fibroids < 12 weeks

Endometrial disease

DUB

Focal pathologies could be missed on imaging, but this

will occur less compared to TVS because of cavity

distension with fluid. A small fibroid encroaching into

the endometrial cavity (submucous fibroid) could be

erroneously considered intramural forming part of a

small fibroid uterus. Cystic looking polyps may be

mistaken for endometrial hyperplasia.


Endometrial Disease

DUB

As for transvaginal ultrasound...

Fibroid uterus >12 weeks size Polyp / SMF

Fibroids < 12 weeks

As for transvaginal ultrasound...


Fibroids < 12 weeks

DUB

As for transvaginal ultrasound..


Fibroids <12 weeks

Endometrial disease

As for transvaginal ultrasound..

Outpatient hysteroscopy

Intrauterine polyp or submucous fibroid Endometrial disease

DUB

Focal pathologies could be missed on imaging, but this

will occur less compared to TVS because of cavity

distension with fluid. OPH cannot visualise the

myometrium and so, in contrast to TVS and SIS,

presence of fibroids cannot be falsely diagnosed. Polyps

may be mistakenly diagnosed as endometrial disease.

137

Table 8.1 continued

True pathology



Endometrial disease

The myometrium is not visualised by OPH and preceding

clinical examination is not sensitive enough to identify

small fibroids. Thus, at OPH a normal cavity would be

found in the presence of the true pathology (small

intramural fibroids). Possible false diagnoses would be

when normal, functional thickened endometrium is

considered falsely to be some form of endometrial

disease (hyperplasia or cancer) or focal lesion (e.g. folds

of normal endometrium mistaken for a polyp).

Fibroid uterus >12 weeks size Polyps / SMF

Endometrial disease

Large fibroids would be missed at OPH because the

myometrium is not visualised, although the majority

would be detected on preceding clinical examination*.

At OPH a normal cavity would be found in the presence

of the true pathology (large intramural fibroids). Possible

false diagnoses would be when normal, functional

thickened endometrium is considered falsely to be

endometrial disease (hyperplasia or cancer) or a focal

lesion (e.g. folds of normal endometrium mistaken for a

polyp).


DUB

A thickened hyperplastic or cancerous endometrium

could appear normal or as a polyp.


Endometrial disease

A normal, thickened endometrium could be considered

falsely to be endometrial disease (hyperplasia or cancer)

or a focal lesion (e.g. folds of normal endometrium

mistaken for a polyp especially in the secretory phase of

the menstrual cycle).

138

Table 8.1 continued

True pathology


Endometrial biopsy

Intrauterine polyp or submucous fibroid Endometrial disease (hyperplasia but not

cancer)

DUB

Endometrial polyp tissue could be mistaken for normal

or complex endometrial hyperplasia but it would be

extremely unlikely to erroneously diagnose a polyp as

endometrial cancer. The focal lesion may be missed by

the biopsy.


Endometrial disease (hyperplasia not cancer)

Cystic pieces of endometrium could be mistaken for

endometrial polyps. Fibroids can distort the uterine

cavity and compact areas of endometrium. If these areas

are sampled they can be mistaken for complex

endometrial hyperplasia

Fibroid uterus >12 weeks size Polyps / SMF

Endometrial disease (hyperplasia not cancer)

As for small fibroids above


DUB

Polyp or DUB were considered the only plausible false

diagnoses

Dysfunctional uterine bleeding Polyp / SMF An endometrial polyp or submucous fibroid were

considered the only possible false diagnoses. It was felt

to be extremely unlikely to mistakenly diagnose any

endometrial disease from a normal sample. *Note that it was assumed that a bimanual gynaecological examination took place in all women with HMB immediately pre-testing. Expert consensus was that this examination would allow the

detection of a large fibroid uterus >12 week. A competent gynaecological examination would normally detect a large fibroid pelvic mass, but in some instances (e.g. poor patient tolerance,

obesity) clinical examination would be less reliable so that the examination would be a false negative. This was assumed to occur in 20% of cases. In the absence of large fibroids, clinical

examination was considered unable to discriminate between any of the other potential pathologies and so examination outcomes were dichotomous; i.e. either large fibroids > 12 weeks or

‘normal’. As regards false positive diagnosis, it was felt that a small fibroid uterus < 12 weeks could be mistaken for large fibroid uterus > 12 weeks, OPH= outpatient hysteroscopy; TVS =

transvaginal scan; SIS= saline infusion sonography; SMF= submucous fibroid; DUB= dysfunctional uterine bleeding.

139

Table 8.2 First and second line treatments for heavy menstrual bleeding according to

underlying diagnosis in women with no desire to retain fertility

Diagnosis Treatment 1 Treatment 2 (only

performed if patient ‘not

satisfied’ with treatment 1)

Endometrial polyp Outpatient polypectomy Levonorgestrel intrauterine

system

Submucous fibroid Transcervical resection of fibroid Levonorgestrel intrauterine

system

Fibroids <12 weeks

size

Levonorgestrel intrauterine

system

Endometrial ablation

Fibroids >12 weeks

size

Total abdominal hysterectomy GP visit

Complex hyperplasia Levonorgestrel intrauterine

system

Total abdominal hysterectomy

Complex hyperplasia

with atypia /

Endometrial Cancer

Total abdominal hysterectomy GP visit

Dysfunctional uterine

bleeding

Levonorgestrel intrauterine

system


GP= general practitioner

Adaptations of the base case tree to assess alternative clinical scenarios

Women being managed during multiple clinic visits

The base case tree was designed to reflect a contemporary ‘one-stop’ clinic to ensure that the

results remain relevant and do not quickly become outdated as services evolve. However, this

approach has not yet been widely adopted across the UK. Traditionally women referred to a

gynaecologist would be reviewed in a general outpatient clinic before any investigation or

treatment was instigated. The patient would have their history taken and be examined and

then the clinician would plan appropriate tests and send the patient away to have these done at

a later date. Weeks or even months later the patient would be seen again in clinic with the

140

results of those investigations and then the clinician would initiate treatment. The model was

adapted to represent the ‘traditional’ multi-stop care model.

In this analysis all tests were performed at separate appointments except for endometrial

biopsy which would be taken at the initial consultant appointment. Transvaginal scan and

saline infusion sonography were assumed to be performed in the ultrasound department at a

later date with the patients having a follow-up appointment to review the results. Outpatient

hysteroscopy required a further consultant appointment for the hysteroscopic assessment. If

polyps were diagnosed, whether by scan or hysteroscopy, patients required a further

hysteroscopy appointment for removal. Submucous fibroids were removed under general

anaesthesia and treatment with the LNG-IUS, hysterectomy and endometrial ablation were

performed as in the base case tree.

Women refractory to LNG-IUS treatment

Alternative analysis was performed by adapting the model to fit with the scenario that all

women referred to secondary care had already received treatment with a LNG-IUS in primary

care but whose symptoms had not resolved. This was to reflect current NICE guidance which

recommends that women receive a LNG-IUS in a primary care setting as first-line treatment

for HMB (142) and only attend secondary care if their symptoms are refractory or structural

abnormality is expected. The prevalence of disease changed within this tree as it was assumed

that patients treated appropriately with the LNG-IUS (DUB, endometrial hyperplasia, fibroids

<12 weeks size) would be less likely to be referred to secondary care than women who were

being treated inappropriately (fibroids > 12 weeks size, polyps, SMF’s, endometrial cancer)

because their symptoms would be more likely to have resolved. Satisfaction rates for

141

treatment of each of the pathologies were used to re-calculate the disease prevalence. LNG-

IUS was no longer a possible treatment within this tree as women had previously failed to

respond to it. The exception to this rule was for women who were dissatisfied following

removal of a polyp or SMF as they now had a ‘normal’ uterine cavity, where as previously

there had been a structural abnormality compromising the effectiveness of the LNG-IUS.

These women received a LNG-IUS as their second treatment following removal of focal

pathology. Endometrial ablation became the first treatment to be offered to women who were

thought to have a ‘normal’ uterine cavity and if this failed, hysterectomy was offered as the

next treatment option. Given that patients already have a LNG-IUS in situ when they attend

secondary care in this scenario the tree was adjusted so that the comparative strategy was ‘no

further treatment’ to represent patients coming to clinic but not having any further treatment

i.e. woman attending the clinic but ultimately deciding to continue with the LNG-IUS.

Women wishing to retain their fertility

The base case analysis was revised to reflect a population who wished to maintain their future

fertility. This meant that endometrial ablation and hysterectomy were no longer possible

treatments, except in the case of endometrial cancer when hysterectomy was still selected as

the treatment of choice. Myomectomy and uterine artery embolisation (UAE) were introduced

as possible treatments in this tree as they are far more likely to be offered to women who wish

to have children than women who have completed their families. Myomectomy was assumed

to be selected over UAE by 80% of women as it is thought to improve fertility to a greater

extent than UAE (201). Following UAE or myomectomy patients who were ‘not satisfied’

with their treatment were offered the other treatment. Hysterectomy was not offered as a

treatment for any benign cause of HMB and women who remained ‘not satisfied’ after two

142

treatments, or after one if no fertility preserving treatment could be offered, received a GP

visit and a new gynaecology outpatient appointment as ‘rescue treatment’.

Clinical data collection

All literature derived data were obtained following systematic searches.

Accuracy data came from published systematic reviews and meta-analyses when possible,

followed by data derived from primary, well conducted test accuracy studies.

For treatment data, systematic quantitative reviews using individual patient data (IPD) were

considered as the highest level of data, followed by systematic reviews of study level data.

Randomised controlled trials were acknowledged as the third step down the hierarchy,

followed by large comparative cohort studies and then uncontrolled observational series.

Prospectively collected data from studies with large populations were considered superior to

small studies and those with retrospectively collected data. When possible data from a purely

premenopausal population were used, however occasionally data came from studies of AUB

incorporating both pre- and postmenopausal women. When possible these data were stratified

by menopausal status.

Disease prevalence

For prevalence of disease underlying HMB symptoms a gold standard test was selected for

confirmation of diagnosis (Table 8.3) A systematic literature review was then performed to

estimate the prevalence of pathologies as estimated by the gold standard confirmatory test.

143

Table 8.3 Gold standard diagnostic tests for uterine pathology

Pathology

Confirmatory test

Polyps


Submucous fibroids


Uterine fibroids < 12 weeks

Pelvic ultrasound

Uterine fibroids > 12 weeks

Pelvic ultrasound

Endometrial disease

Histological sampling


Diagnosis of exclusion

The searches for prevalence are reported in Appendix 7.1-7.4. As dysfunctional uterine

bleeding is a diagnosis of exclusion a diagnostic test was not used as one of the search terms.

The prevalence of each diagnosis was determined from published studies, using systematic

reviews when possible. Seven quality criteria were used to assess the quality of identified

studies (see Table 8.4) with a score from one to three being awarded for each criterion giving

a maximum score of 21.Data from the highest scoring papers were used to populate the

decision model.

Table 8.4 Quality criteria for disease prevalence studies

Criterion Points awarded

3 2 1

Data collection

Prospective Retrospective Not clear

Consecutive patients

Yes No Not reported

Population size

>500 100-500 <100

Menopausal status Premenopausal Mixed but >50%

premenopausal

Postmenopausal

Data collection All have the gold

standard test

Selection prior to

gold standard test

Inferior test

Proportion having the

gold standard test

>90% <90%

Pathology clearly defined

Clear definition Unclear definition No definition

144

Data regarding the prevalence of fibroids (intramural and subserosal) were taken from a

database of 473 women with HMB, held at Birmingham Women’s Hospital (BWH) in the

absence of a better quality published data set.

Data regarding the prevalence of DUB were often not specified within studies in contrast to

organic pathologies. As the overall prevalence of disease must add up to 1 (100%) in the

economic model it was decided that the prevalence of DUB (which is a diagnosis of

exclusion) would be altered to become the remaining proportion once the prevalence of other

pathologies had been estimated. The impact of this manipulation was tested with sensitivity

analyses.

Test success and accuracy

To identify diagnostic data regarding the feasibility and accuracy of the tests under evaluation

broad search terms were used for the literature searches to ensure a high level of sensitivity. If

searches retrieved a large number of studies they were restricted to review articles only. In the

case of searches for SIS, the search was qualified by the population under scrutiny because

preliminary, broad searches were retrieving a large number of articles which evaluated the test

in postmenopausal women and women with infertility. Duplicate articles were removed, the

abstracts of all remaining articles were read and the full text of relevant papers retrieved based

upon the following selection criteria:

Population Heavy menstrual bleeding

Intervention Outpatient test to evaluate the uterus (TVS, SIS, EBx, OPH)

Outcome Feasibility (success rate) or test accuracy for uterine pathology

Study design Restricted to systematic reviews if available

145

Although systematic reviews report high quality aggregated data of feasibility and/or test

accuracy, the data reported did not always give the relevant accuracy data for all of the

pathologies. In these circumstances restriction on study design was removed and in the

absence of relevant primary study data, additional searches to look for data regarding the

accuracy of the test without regard for population characteristics (heavy menstrual bleeding)

were undertaken as were searches using specific pathologies as the population of interest. The

quality of the studies was assessed using the criteria in Table 8.5. The highest scoring studies

were selected. If studies were of equal quality the largest was selected and if this did not

discriminate, blinding and interval between tests were taken into consideration.

Table 8.5 Quality criteria for assessing test accuracy studies

Criterion

Explanation

Population size

>100 women

Type of bleeding

>75% with heavy menstrual bleeding

Menopausal status

>70% premenopausal

Data collection

Prospective

Reference test

Appropriate ‘gold’ standard test applied (see Table 8.3)

Blinding Present (diagnostic test and the gold standard test are

performed by different, blinded clinicians).

Interval between tests

Within the same menstrual cycle

Cross tabulation

Data presented in a 2x2 table

Total 8 points max (one point awarded for each criteria present)

Details of searches are given in Appendix 7.5-7.8. An unsuccessful test was defined as failure

of the test to provide a diagnosis. This may arise for a number of reasons such as an inability

to pass the ultrasound probe (TVS, SIS) into the vagina, instrument the uterine cavity (SIS,

146

EBx, OPH) or because of inadequate visualisation (TVS,SIS, OPH). In the case of EBx,

failure also included a successfully completed test but the subsequent finding of a tissue

sample insufficient for diagnosis.

Accuracy data tended to be reported as sensitivity and specificity along with their respective

95% confidence intervals. These true positive and true negative rates were then used to

calculate the false positive and negative rates (see Table 8.6).

Table 8.6 Accuracy value calculation using sensitivity and specificity

Accuracy value Calculation

True positive rate Sensitivity

True negative rate Specificity

False positive rate 1 – specificity

False negative rate 1 – sensitivity

When data were reported as likelihood ratios, sensitivity and specificity values were derived.

Sensitivity and specificity along with their 95% confidence intervals were calculated from

raw data when necessary. In order to simplify the decision model, underlying pathologies

were grouped when appropriate i.e. similar aetiologies and/or commonly associated

treatments (see assumptions section). One problem with this approach was that accuracy data

for these categories were often reported separately for each respective pathology. Therefore

the figures were combined, weighting them according to the proportion of the pathology

category they made up.

147

Taking EBx as an example test:

Endometrial hyperplasia = 60% of the endometrial disease category

Atypical endometrial hyperplasia / cancer = 40% of the endometrial disease category

The sensitivity value for endometrial biopsy for hyperplasia is 0.81

The sensitivity value for endometrial biopsy for atypia/cancer is 0.86

Combined sensitivity is (0.6 x 0.81) + (0.4 x 0.86)

Sensitivity = 0.83

A final data manipulation was needed when tests were reported in the literature as having a

sensitivity or specificity of 1. These perfect accuracy data were rounded down to 0.99 as

although tests can have high predictive values, no test can be reported as completely accurate.

Treatment satisfaction data

Systematic searches of the literature were conducted to identify patient satisfaction data at one

year post-treatment. The electronic bibliographic databases EMBASE and Medline were

searched using search terms for the relevant treatment, combined with menorrhagia or HMB,

and satisfaction along with their alternatives. Searches and outputs are detailed in Appendix

7.9-7.13.

For the purpose of the model, effectiveness data in terms of patient satisfaction were needed

according to underlying pathology. Whilst treatment outcome data are reported for women

with HMB, in some cases the underlying diagnoses were not ascertained or treatments were

not utilised when contraindicated. However, the diagnostic model required data for treatment

outcomes not only used appropriately but also for when false diagnoses were made on testing

148

and so suboptimal or inappropriate treatments would be applied. When this data could not be

identified in the published literature, the panel of gynaecologists were surveyed for their

opinions regarding the likely treatment efficacy. The median values obtained in this way were

used as the satisfaction rate and the range of values were used for sensitivity analysis.

Systematic reviews with meta-analysis were considered the highest quality data, followed by

randomised controlled trials. When 12 month data weren’t available the data reported closest

to twelve months were used. If the time of reporting was not stated within the study the data

were only used when no other appropriate data had been identified. For endometrial ablation,

data were used that exclusively examined second generation techniques (6) as first generation

methods are now seldom used.

Results of clinical data collection

Disease prevalence

Uterine polyps

The searches for prevalence of endometrial polyps identified 845 studies (Appendix 7.1).

Seven studies were selected and the manuscripts obtained for further analysis. The majority of

studies were rejected because they did not report prevalence or referred to postmenopausal or

infertile populations. The highest quality study (score 20/21) reported a uterine polyp

prevalence of 18% (202). The range in prevalence was wide however, with the next two best

studies reporting a prevalence of 3.7% (203) and 33.9% (204).

149

Submucous fibroids

Searches for the prevalence of fibroids in women with heavy menstrual bleeding (Appendix

7.2) identified 134 papers, 3 of which were selected and assessed for quality and the

prevalence data were extracted. Papers were discarded if they did not report prevalence of

submucous fibroids in premenopausal women and if the reference test used was not the

selected gold standard. The two highest scoring papers reported differing values of 21.9% and

7.4% for the prevalence of submucous fibroids (202;204) with 21.9% coming from the best

paper (202) (quality score 18/21).

The prevalence of polyps and submucous fibroids reported in the highest quality studies were

intended for use in the decision tree. However, whilst examining papers that reported the

accuracy of hysteroscopy a systematic review of diagnostic hysteroscopy was identified

which reported the prevalence of polyps and submucous fibroids (205). This systematic

review meta-analysed over 3000 procedures. The prevalence of polyps was reported as 21%

and that of submucous fibroids as 25%. However, polyps and submucous fibroids are

estimated to co-exist in approximately one third of women (206) and so the value for

submucous fibroids was reduced by one third to 17% to account for this. Thus the derived

prevalence rate used for the pathology category ‘endometrial polyps/SMF’ was 38% (0.38).

Fibroids

Two studies were identified that looked at the prevalence of intramural or subserosal fibroids

out of 134 studies identified from the original search (Appendix 7.2). One study only

contained 80 participants (206) and the second was a study of women being scanned for a

variety of symptoms (pain, worry, AUB, suspected fibroids) and not just heavy menstrual

bleeding (207). The prevalence values reported by the two studies were 57.7% and 23.5%

150

respectively. Neither study specified whether submucosal fibroids were excluded and the

patient populations were heterogeneous, preventing meta-analysis. Both studies scored 17 for

quality but in view of the small size of the first study (80 patients), the poorly defined

population in the second study and the large discrepancy in the reported values, a database of

473 women held at the BWH was consulted. This database showed that 19% (88/473) women

had intramural fibroids less than 12 weeks size and that 6% (28/473) had fibroids greater than

12 weeks size, thus these data were used as the prevalence values within the decision tree.

Endometrial disease

The searches for prevalence of endometrial hyperplasia (Appendix 7.3) diagnosed by

histology samples identified 86 studies of which five were chosen for review. Two studies

were high quality (both scoring 20/21) and reported similar values for the prevalence of

endometrial hyperplasia without atypia (3.0% and 2.4%). One study was a retrospective

review of histology samples (208) and the second a retrospective audit of hysteroscopy

findings (202). The same two studies were also identified as the best from the searches for

studies reporting the prevalence of endometrial cancer (374 studies identified, 4 assessed

further) (Appendix 7.4). Both studies were large and contained only premenopausal women.

Meta-analysis of these two studies was performed to calculate the prevalence data more

precisely. Data for endometrial hyperplasia without atypia were analysed separately from data

for endometrial hyperplasia with atypia and endometrial cancer because the clinical

implications of diagnosis and optimal therapeutic interventions differ. Meta-analysis was

performed by converting the values from the studies to log odds and standard errors using

Excel (Microsoft) this data was then copied across to RevMan (Cochrane Library) and

analysed as the generic inverse variance using random effects analysis. The output from

RevMan is the odds, so these were copied back into Excel and converted to values to give

151

prevalence rates with 95% confidence intervals. The value calculated for prevalence of

endometrial hyperplasia without atypia was 3% and that for endometrial hyperplasia with

atypia or endometrial cancer as 2%. Therefore the value for ‘endometrial disease’ was the

sum of the two, 5%.

Dysfunctional Uterine Bleeding

In all decision trees the sum of all the branches from a common stem must add to 1. It was

inevitable that the prevalences for the five pathology groups would not add up to one as the

data came from a variety of sources and so dysfunctional bleeding was used as the ‘buffer’ to

make the prevalences add to one (given DUB is a diagnosis assigned after exclusion of other

identifiable ‘organic’ pathologies). This resulted in the prevalence of DUB being set at 32%.

A summary of the derived disease prevalences is given in Table 8.7.

Table 8.7 Estimated prevalence of pathologies in heavy menstrual bleeding: Base case

Disease Prevalence Sensitivity analysis

(range)

Source

Polyps/Submucous fibroids

0.38 0.2-0.5 (EP

estimate)

SR (205)

Fibroids <12/40

0.19 0.15-0.22†

BWH database

Fibroids >12/40

0.06 0.04-0.08†

BWH database

Complex endometrial

hyperplasia

0.03 0.02-0.03†

Meta-analysis of two

studies (202;208)

Atypical hyperplasia / cancer

0.02 0.01-0.02†


studies (202;208)

Endometrial disease (all

hyperplasias and cancer)

0.05 0.03-0.05†


studies (202;208)


0.32 Remaining

proportion

Remaining

proportion used so

that total sums to 1 SR= systematic review; †= 95% confidence interval; EP= expert panel

152

To calculate the prevalence of disease within the decision tree for women already treated with

a LNG-IUS in primary care, the data were weighted by taking the proportion of women who

would not be satisfied with a LNG-IUS for each pathology (see rates later on in this chapter)

and dividing each by the total, so that the sum of them all came to one and became the new

prevalences for the decision tree (Table 8.8).

Table 8.8 Recalculated disease prevalence for HMB in women refractory to treatment

with a LNG-IUS

Disease Original

prevalence

(a)

Proportion

dissatisfied

with LNG-IUS

(b)

Proportion

coming to

gynaecology

(a x b)

(c)

New prevalence

(c / total c)

Polyps/Submucous

fibroid

0.38 0.60 0.228 0.592

Fibroids <12/40

0.19 0.17 0.032 0.083

Fibroids >12/40

0.06 0.71 0.043 0.111

Endometrial disease

0.05 0.56 0.028 0.073

Dysfunctional uterine

bleeding

0.32 0.17 0.054 0.141

Total

0.3853 1.0 LNG-IUS= levonorgestrel intrauterine system

Test success

Success data for each of the four separate tests came from systematic reviews and meta-

analysis (see Table 8.9).


The search strategy for outpatient hysteroscopy identified three systematic reviews

(198;199;205) from 1095 studies. Two studies reported test success (205) (198). The

outpatient hysteroscopy value came from one of the systematic reviews (205) of diagnostic

153

hysteroscopy. This quantitative systematic review reported success data separately for all

women having outpatient hysteroscopy and also for premenopausal women. However, the

data for OPH included some postmenopausal women and the data for premenopausal women

contained some women undergoing hysteroscopy as an inpatient under general anaesthesia. In

total 2643/3158 (84%) procedures were done as outpatient hysteroscopies and only 306

women were specified as being postmenopausal, therefore it was decided to use the success

rate value for outpatient hysteroscopy (0.97). This value was supported by the second review

of predominantly pre-menopausal women (71%) but included both inpatient and outpatient

hysteroscopy. The reported success rate in this meta-analysis was similar at 95.8% (CI 95.5-

96.1%). (198).


There were no systematic reviews of premenopausal women undergoing transvaginal

ultrasound that reported test success rate. Only one small study of 43 women reported the

success rate of TVS and this was 100% (209). A previously used systematic review of women

with postmenopausal bleeding reported a mean success rate of 100% with a standard

deviation of 2% when data from 16 studies were meta-analysed (210). The value 0.99 was

used because no test is 100% successful; women do occasionally refuse to have the test,

cannot tolerate it or the visualisation on imaging is too poor to make a diagnosis.


Searches for reviews of saline infusion sonography identified 257 studies, two of which were

systematic reviews (197;199) which were selected for data extraction. The first systematic

review included meta-analysis of the accuracy of saline infusion sonography in a population

of women with abnormal uterine bleeding (>50% premenopausal) and reported that the

154

success rate of the test in premenopausal women was 94.8% with confidence intervals of 93.5

-96.1% (197). A similar success rate of 95% reported by the second review examining SIS in

premenopausal women supported this value (199).

Endometrial biopsy

Searches for reviews of endometrial biopsy identified two systematic reviews, one for the

diagnosis of endometrial hyperplasia (150) and one for the diagnosis of endometrial cancer

(151). The latter review of endometrial biopsy for diagnosis of endometrial cancer had a

mainly postmenopausal population. Therefore the study looking at endometrial hyperplasia

where the majority of women were premenopausal (50% were known to be premenopausal,

25% were known to be postmenopausal and 25% unknown) was used. This study found that

76/881 (8.6%) tests failed or were insufficient for histological diagnosis (150) and so a failure

rate of 9% (0.09) was used. The raw data were used to calculate confidence intervals.

Test combinations

When combinations of tests were performed, the success rates of the individual tests were

multiplied within the tree to calculate the success rate. The success rates of combined tests

will always be worse than tests performed individually which is reflected by this

manipulation. However it does not take into account whether the failure of one test is

dependent upon the next, for example, if a hysteroscopy fails because of a stenosed cervical

canal, an endometrial biopsy would be very likely to also fail.

155

Table 8.9 Diagnostic test success rate data

Test Success rate Sensitivity analysis

(range)

Source

Outpatient hysteroscopy 0.97 SR(205)


scan

0.99* SR (210)

Endometrial pipelle biopsy 0.91 0.89- 0.93† SR (150)

Saline infusion sonogram 0.95 0.94-0.96† SR(197)

*reported as 0 but reduced to 0.99 as unlikely. †= 95% confidence interval, SR= systematic review

Test accuracy


The accuracy data for hysteroscopy came from the two systematic reviews (see search in

Appendix 7.5) used for test success and were identified by the outpatient hysteroscopy

database searches (198;205). The first review looked at the accuracy of hysteroscopy for

diagnosing intrauterine abnormalities in women presenting with abnormal uterine bleeding

(pre and postmenopausal) using histopathology specimens as the reference standard. This

study provided the data for polyps, submucous fibroids and dysfunctional uterine bleeding.

This study has limitations as it has a mixed population and only 84% of procedures are

specified as outpatient hysteroscopies. However, no other large data sets reporting accuracy

exist. Data for the sensitivity and specificity of polyps (0.94 and 0.92) and submucous

fibroids (0.87 and 0.95) are clearly reported in the paper and were combined to values for the

combination group of polyps/submucous fibroids. Studies included in this large, systematic

quantitative review of hysteroscopy report the accuracy of a test for diagnosing pathology

rather than a normal cavity. Thus data for the accuracy of outpatient hysteroscopy in

diagnosing dysfunctional uterine bleeding could not be identified, a reflection of the fact that

DUB is considered a diagnosis of exclusion. Therefore the data values from the review were

reversed i.e. used the proportion not diagnosed as abnormal and assigned these women as

156

DUB. In practice this results in the sensitivity and specificity being reversed. The sensitivity

and specificity reported for diagnosing abnormalities were 0.94 and 0.89 respectively, so for

diagnosing ‘no’ abnormality the values reverse and the sensitivity becomes 0.89 and

specificity 0.94.

The data regarding the accuracy of outpatient hysteroscopy for diagnosis of endometrial

disease comes from the second identified review (198) which specifically looks at this

question in a mixed population with 29% of the women specified as postmenopausal. This

large systematic review meta-analysed data from 65 studies (26,346 women) that compared

outpatient hysteroscopy to endometrial histology results and reported that the sensitivity for

diagnosis of endometrial disease (cancer and hyperplasia) was 0.78 and that specificity was

0.96, hence these values were used in the decision tree.


The search for transvaginal ultrasound accuracy data included terms for abnormal uterine

bleeding and yielded 420 studies once duplicates had been removed (Appendix 7.6). Thirty-

seven studies were selected for further assessment regarding accuracy of transvaginal scan.

Only one of the selected papers was a systematic review (199) but as the studies included

were heterogeneous, no meta-analysis was performed and no useful data could be extracted.

The thirty-seven studies identified were assessed for quality and accuracy data were extracted

for the different pathologies.

A prospective comparative study was selected as the highest quality paper reporting the

accuracy of TVS for diagnosing polyps and submucous fibroids (211). The study compared

the TVS diagnosis to hysteroscopic diagnosis. The tests were performed by different

157

clinicians and each one was blinded to the other result. The tests were performed within 24

hours of each other and the majority of women included in the study were premenopausal.

The sensitivity and specificity of TVS for diagnosing polyps and submucous fibroids were

reported as 0.45 (95% CI 0.32-0.58) and 0.78 (0.62-0.89), and these data were used to

populate the decision tree. Data from the same study were used for accuracy of TVS for

diagnosing endometrial disease reported as sensitivity 0.57 (95% CI 0.19-0.90) and specificity

0.66 (0.55-0.76).

Only one study was identified which reported the accuracy of TVS for diagnosing intramural

or subserosal fibroids. The study aimed to assess the accuracy of TVS for diagnosing

adenomyomas and fibroids and its ability to distinguish between the two pathologies by

comparing the scan results to hysterectomy specimens (212). The mean age of women

included in the study was 46.7 (range 35.7-51.8) years and 172 of the 206 women had

menorrhagia or dysmenorrhoea. The sensitivity and specificity of TVS for diagnosing fibroids

were reported as 95.1% and of 82.0% and were used as the accuracy values for TVS diagnosis

of fibroids.

As with outpatient hysteroscopy, no studies reported the accuracy of TVS for diagnosing a

normal uterus, so this had to be derived indirectly. The largest high quality study (7 points)

that reported the accuracy of TVS for diagnosing abnormality (213) was selected and the

sensitivity and specificity were reversed. This study evaluated 770 women with HMB to

establish the accuracy of TVS for diagnosing a composite of all pathologies labelled

‘intrauterine disease’ by comparing the scan results to the results of hysteroscopy. As the

reported sensitivity and specificity for abnormality were 0.96 (95% CI 0.934 to 0.972) and

158

0.86 (95% CI 0.823 to 0.898) they were reversed and 0.86 (95% CI 0.823 to 0.898) was used

as the sensitivity and 0.96 (95% CI 0.934 to 0.972) as the specificity.


Searches for the accuracy of saline infusion scan identified 157 studies of saline scan for

menorrhagia (Appendix 7.7). Forty-one papers were selected and assessed for quality and any

reported accuracy data were extracted. Two systematic reviews were identified (199) (197),

one of which was the study with no meta-analysis (199) identified during the TVS searches.

The second was a systematic review and meta-analysis of diagnostic studies that compared

SIS to either hysteroscopy or histopathology obtained at hysteroscopy or hysterectomy and

reported the accuracy of the test for diagnosing intrauterine abnormalities (197). The analysis

included twenty-four studies and more than 50% of the population were premenopausal. The

accuracy of saline infusion sonography for diagnosing endometrial polyps and submucous

fibroids was reported as a secondary outcome after meta-analysis of fifteen homogenous

studies. These values were combined and weighted to calculate a value for the two

pathologies combined. Once again data for a normal cavity were not available and so the

sensitivity and specificity data for abnormalities were reversed.

The accuracy of SIS for diagnosing endometrial disease was not reported in the systematic

reviews. Of the thirty-nine remaining studies, four reported the accuracy of SIS for diagnosis

of endometrial disease (hyperplasia and or cancer). When the quality was assessed, two of the

studies scored 6 points (211;214) but the values that they reported for sensitivity were very

different with one reporting a sensitivity for endometrial hyperplasia as 0.94 (214) and the

second reporting sensitivity for endometrial hyperplasia and cancer as 0.29 (211). The study

sizes were very similar, as were the proportions of premenopausal women, however one of

159

the studies had a much higher prevalence of endometrial disease (17%) than would be

expected in a mainly premenopausal population (214) and the interval between the SIS and

the reference test was up to fourteen days when in the second study it was just 24 hours (211).

The first study (214) used a cut-off threshold for women of reproductive age of 8mm for

diagnosing abnormality. Using an endometrial thickness cut-off level is crude and more

applicable to a postmenopausal population where endometrial thickness is constant rather than

changeable according to the menstrual cycle In contrast, the second study diagnosed

abnormalities based upon clinical features seen at SIS rather than defining abnormalities

(211). The culmination of these factors resulted in the decision to use data from the second

study (211) to populate the decision tree.

Endometrial biopsy

Searches for accuracy data of endometrial biopsy identified four studies with data regarding

accuracy of the test. One of the studies was an RCT which looked at the use of three

diagnostic tests, including endometrial biopsy, in groups of women at a specified risk of

endometrial cancer (155). The population included pre and postmenopausal women and only

reported the accuracy of pipelle for diagnosing endometrial cancer. Two of the remaining

selected studies were systematic reviews one of which looked at diagnosis of endometrial

hyperplasia (150) and a second which looked at endometrial cancer (151). Both studies used

histopathology samples as the reference standard. These studies were limited in that they had

mixed populations of pre and postmenopausal women. As postmenopausal women have an

atrophic endometrium, focal lesion are more likely to be sampled and thus the effect of the

postmenopausal women within these three studies may increase the sensitivity of the test

above what may be expected in a purely premenopausal population. As systematic reviews

with meta-analysis were rated as higher quality evidence than RCT data the values from the

160

systematic reviews (150;151) were used for accuracy of diagnosing endometrial disease. In

these papers, likelihood ratios were reported, but in order to populate the model, the

sensitivities and specificities were used which had been reported in the thesis from which the

papers were taken (Clark TJ 2004; PERSONAL COMMUNICATION).

The fourth study reported the accuracy of endometrial biopsy for diagnosis of endometrial

polyps (215). One hundred and seventy-six consecutive patients (77% premenopausal) who

were scheduled for dilatation and curettage underwent transvaginal ultrasound and

endometrial pipelle® biopsy prior to their surgery. The biopsy samples and the curettings

were examined by different pathologists who were blinded to the other result. The paper

reports these results in a cross-tabulated fashion enabling calculation of the sensitivity and

specificity of endometrial biopsy for the diagnosis of endometrial polyps (216) as 0.997 (95%

CI 0.973-1) and 0.003 (95% CI 0-0.027) respectively. These values were converted to the true

positive and false positive values which were used in the decision tree. The some cross

tabulation data were used to calculate the accuracy for diagnosis of dysfunctional uterine

bleeding by looking at the accuracy for benign endometrium (excluding endometrial polyps

and hyperplasia). The sensitivity was calculated as 0.953 (95% CI 0.895 to 0.98) and

specificity as 0.971 (95% CI 0.902 to 0.992) and the corresponding true and false positive

values used accordingly.

No studies reported the use of endometrial biopsy for diagnosis of submucous fibroids so

further searches were performed that did not include ‘heavy menstrual bleeding‘ or its

associated terms. One study of 330 postmenopausal women was identified that reported the

accuracy of endometrial biopsy for diagnosis of submucous fibroids. This prospective study

161

compared Novak catheter samples to histopathology samples obtained during surgery to

establish accuracy of the blind biopsy. The reported sensitivity and specificity of the Novak

catheter were 13% and 100%. The specificity was reduced to 99% (no ‘perfect’ values were

deemed to be plausible) and the values were converted accordingly for use in the decision

tree.

The accuracy estimates, along with ranges for use in sensitivity analyses and data sources are

summarised in Table 8.10.

Table 8.10 Test sensitivity for specific pathologies

Variable Baseline

sensitivity

Sensitivity

analysis (range)

Source

OPH polyps 0.94 0.92-0.96† SR (205)

OPH submucous fibroids 0.87 0.81-0.92† SR (205)

OPH polyps / SMF 0.91 0.87-0.94† Composite of polyp and

SMF values

OPH endometrial disease 0.78 0.76-0.80† SR (198)

OPH DUB 0.89 0.87-0.90† SR (205) values reversed

so for no pathology

TVS polyps / SMF 0.45 0.32-0.58† Prospective comparative

study(211)

TVS intramural fibroids 0.95 Prospective

observational(212)

TVS endometrial disease 0.57 0.19-0.90† Prospective comparative

study (211)

TVS DUB 0.86 0.82- 0.90† Prospective observational

study (213)values

reversed so for no

pathology

162

Table 8.10 continued

Variable Baseline

sensitivity

Sensitivity

analysis (range)

Source

EBx polyps 0.41 0.14-0.76†

Observational

prospective cohort (215)

EBx SMF 0.13 - Prospective comparative

study (217) NB PMB

EBx polyp / SMF 0.27 0.14-0.76 † Composite of polyp and

SMF values

EBx endometrial

hyperplasia

0.66 0.47-0.81† SR (150)

EBx cancer / atypia 0.94 0.84-0.99† SR (151)

EBx endometrial disease 0.78 0.62-0.88† Composite of hyperplasia

and cancer / atypia values

EBx DUB 0.95 0.90-0.98† Observational

prospective cohort (215)

2x2 created and benign

data used to calculate

SIS polyps 0.86 0.81-0.91† SR (197)

SIS SMF 0.87 0.79-0.92† SR (197)

SIS polyps / SMF 0.87 0.80-0.92† Composite of polyp and

SMF values

SIS endometrial disease 0.29 0.05–0.71† Prospective comparative

study(211)

SIS DUB 0.88 0.85-0.92† SR (197) values reversed

so for no pathology

OPH= outpatient hysteroscopy; TVS = transvaginal scan; EBx= endometrial biopsy; SIS= saline infusion sonography; SMF=

submucous fibroid; DUB= dysfunctional uterine bleeding; SR= systematic review; † = 95% confidence interval

False positive rates

Although the false positive rates were calculated from the specificity data (1-specificity =

false positive rate) the derived values could not always be used in their pure form within the

163

decision model. This is because where a false diagnosis is made there are several possible

erroneous options. However, the branches from one stem in the tree have to sum to one. This

meant that all of the possible false diagnoses needed false positive rates which added up to

one. To overcome this, if one of the erroneous diagnoses was DUB (i.e. normal) the FPR’s for

the other pathologies would be used and the value remaining would become the FPR for

DUB. When DUB was not a possible false diagnosis each FPR was divided by the sum of

them together to weight the respective values appropriately:

Example 1:- 1.If it is assumed that for TVS the possible false diagnoses were

polyp/SMF (FPR=0.22), endometrial disease (FPR=0.34) and normal (FPR = 0.04).

The only value which changes is the value for normal (i.e. DUB) which becomes 1-

(0.34+0.22) = 1-0.56 = 0.44.

Example 2:- If it is assumed that for TVS the possible false diagnoses are polyp/SMF

(FPR= 0.22) or fibroids <12 weeks size (FPR= 0.18). Both values are then divided by

the combined FPRs [(0.22+0.18) = 0.40] to weight them so that the sum of the two

values equals one e.g. 0.22/0.40= 0.55 and 0.18/0.40= 0.45.

This rule was used consistently throughout the decision tree. Weighting the values meant that

the reported confidence intervals could not be used in subsequent sensitivity analyses so beta

distributions were used to calculate appropriate confidence intervals. Table 8.11 details the

false positive rates that were used within the tree as well as explanations as to how they were

derived.

164

Table 8.11 Manipulation of false positive rates and their use within the decision tree Test True diagnosis False diagnosis Value Why

TVS Polyp/SMF Fibroids <12wk 0.18 TVS FPR for intramural fibroids

Endometrial disease 0.34 TVS FPR for endometrial disease

Normal 0.48 Remaining

Fibroids <12wk Polyp/SMF 0.22 TVS FPR for polyp / SMF

Endometrial disease 0.34 TVS FPR for endometrial disease


Fibroids >12wk Polyp/SMF 0.55 Unconditional FPR s weighted to add to 1

Fibroids <12wk 0.45

Endometrial disease Polyp/ SMF 0.22 TVS FPR for polyp / SMF

Fibroids <12wk 0.18 TVS FPR for Intramural fibroids


DUB Polyp /SMF 0.30 Unconditional FPR s weighted to add to 1

Fibroids <12wk 0.24

Endometrial disease 0.46

OPH Polyp / SMF Endometrial disease 0.04 OPH FPR for endometrial disease


Fibroids <12wk Polyp/SMF 0.6 Unconditional FPR s weighted to add to 1*




Endometrial disease Polyp/SMF 0.06 OPH FPR for polyp/SMF


DUB Polyp/SMF 0.6 FPR s weighted*


165

Table 8.11 continued

Test True diagnosis False diagnosis Value Why

EBx Polyp/SMF Normal 0.95 Remaining

Comp hyp 0.05 EBx FPR comp hyp

Fibroids <12wk Polyp/SMF 0.038 Unconditional FPR s weighted to add to 1

Comp hyp 0.962


Comp hyp 0.962

Endometrial disease Polyp/SMF 0.002 EBx FPR for polyp/SMF


DUB Polyp 1 As no alternative disease

SIS Polyp/SMF Fibroids <12wk 0.18 TVS FPR for intramural fibroids

Endometrial disease 0.02 SIS FPR for endometrial disease


Fibroids <12wk Polyp/SMF 0.13 SIS FPR for polyp/SMF

Endometrial disease 0.02 SIS FPR for endometrial disease



Fibroids <12wk 0.58

Endometrial disease Polyp/SMF 0.13 SIS FPR for polyp/SMF

Fibroids <12wk 0.18 TVS FPR for intra fib


DUB Polyp/SMF 0.39 Unconditional FPR s weighted to add to 1

Fibroids <12wk 0.55

Endometrial disease 0.06 OPH= outpatient hysteroscopy; TVS = transvaginal scan; EBx= endometrial biopsy; SIS= saline infusion sonography; SMF= submucous fibroid; DUB= dysfunctional uterine bleeding;

SR= systematic review; FPR= false positive rate

166

Test combinations

As the test combination trees display tests in series, no new values needed to be calculated for

data accuracy for the combination trees.

Treatment satisfaction data

Levonorgestrel intrauterine device

The medical database searches identified 2987 studies using the terms ‘levonorgestrel

intrauterine device’, ‘heavy menstrual bleeding’ and their associated phrases. Eighty-two

studies reported data regarding the effectiveness of the LNG-IUS. They were selected based

on whether they reported effectiveness data for the LNG-IUS in any form for any pathology.

The highest quality data for each pathology were then selected and used within the analysis

(Appendix 7.9). The LNG-IUS works optimally when used to treat DUB and so data that

reported patient satisfaction when used in women with DUB as their underlying pathology

were identified and used as a reference. One of the selected studies was a systematic review

with individual patient data (IPD) meta-analysis which looked at the relative effectiveness of

hysterectomy, endometrial destruction and levonorgestrel intrauterine devices (LNG-IUS)

(184). The review used 12 month follow up data to report rates of dissatisfaction when

comparing the different treatments. The dissatisfaction reported for LNG-IUS overall at 12

months was 17% (22/128). This was converted to a satisfaction rate of 83% and confidence

intervals were calculated from the data so that the values used for satisfaction with LNG-IUS

when used to treat DUB were 0.83 (95% CI 0.76-0.89).

No suitable studies were identified that reported the satisfaction level of premenopausal

women with polyps, fibroids, endometrial hyperplasia or cancer if they were treated with an

LNG-IUS. Any studies reporting this outcome were either very small (< 50 patients) or had a

167

mainly postmenopausal population. One study reported predictors of outcome for LNG-IUS

and stated that small fibroids were not predictive of outcome (161), therefore an extrapolation

was made from this and the same date were used for fibroids less than twelve weeks size as

was used for DUB from the IPD study(184). For polyps / SMFs and for fibroids greater than

twelve weeks size no studies were identified that reported patient satisfaction associated with

the use of the LNG-IUS so the panel of gynaecologists were asked to estimate how effective

the device would be in women with theses pathologies. The values of presumed treatment

satisfaction for polyps/SMFs were inconsistent ranging from 20% to 85% with a median

value of 40%. For fibroids > 12 weeks size the range was similarly imprecise varying from

10% to 75% with a median value of 29%. The median values formed the point estimates and

the ranges were used in the sensitivity analysis.

One study was identified that reported the regression of endometrial disease with the use of

the LNG-IUS (162), however only 37 of the women were premenopausal. The study reported

that at 12 months 69/80 (86%) with complex hyperplasia had regressed and that 6/9 (66%)

with atypical hyperplasia had regressed. This study was reported from data collected at BWH

in 2008, however the database continued to be updated and so up-to-date data were used to

produce values for endometrial hyperplasia and cancer. The database records follow-up data

for women who have been diagnosed with endometrial hyperplasia with and without atypia,

who are being treated with systemic or local progestogens. Women who were being treated

with a LNG-IUS for endometrial disease were identified and their 6 and 12 month follow-up

data examined. If they were still using the LNG-IUS at 12 months it was assumed that they

were satisfied with it. If they had undergone hysterectomy or had the IUS removed they were

counted as unsatisfied. One-hundred and one premenopausal women, 95 with complex

168

hyperplasia and 6 with complex hyperplasia with atypia were identified in the database.

Thirteen of the women with complex hyperplasia had undergone a hysterectomy before 12

months so 82/95 (86%) women were considered satisfied. This value is consistent with the

original study (162). However this treatment success rate is greater than the aforementioned

literature derived estimates for successful treatment outcomes in women without uterine

pathology (DUB). This was considered highly unlikely so the value for the satisfaction rate of

treatment with LNG-IUS in women with endometrial hyperplasia was reduced to 83% to

make it the same value as for the DUB population. Confidence intervals were calculated for

the original data but they were varied around a point estimate of 0.83 instead of 0.86. Of the

six women who were treated for complex endometrial hyperplasia with atypia, three had

undergone hysterectomy by twelve months, giving a satisfaction rate of 50%.

It was assumed that the patients whose underlying disease was endometrial cancer would not

be satisfied with the LNG-IUS treatment as the inappropriate treatment would mean

persistence or worsening of their symptoms. In the database the ratio of pre-menopausal

women with atypical endometrial hyperplasia to premenopausal women with endometrial

cancer was 0.59: 0.41 so each was multiplied by the satisfaction value to produce a composite

‘satisfaction’ value for the group ‘atypia/cancer’:

Atypical hyperplasia satisfaction rate= 0.50 prevalence= 0.59

Endometrial cancer satisfaction rate= 0.00 prevalence= 0.41

Overall composite satisfaction rate is (0.50 x 0.59) + (0 x 0.41) = 0.295

Value used for treatment satisfaction = 0.3

169

Similarly, the value for complex hyperplasia was used proportionally with the value for

‘atypia/cancer’ to produce an overall satisfaction value for ‘endometrial disease’.

Complex hyperplasia satisfaction rate= 0.83 prevalence= 0.60

Atypia/cancer satisfaction rate= 0.30 prevalence= 0.40

Overall composite satisfaction rate (0.60 x 0.83) + (0.30 x 0.40) = 0.44

Value used for treatment satisfaction = 0.44


To populate the decision tree, values were needed for satisfaction after endometrial ablation

for DUB and for fibroid uteri. As endometrial ablation requires women to undergo an

endometrial biopsy and either a TVS, SIS or OPH prior to the procedure it was assumed that

any intrauterine pathology would be picked up by one of these pre-ablation tests and then

treated appropriately. Thus outcome data for polyps/SMF and endometrial disease were not

required. Searches for endometrial ablation identified 319 relevant studies, 8 of which were

systematic reviews, with 4 containing meta-analysis (184;193;194;218) (Appendix 7.10a-

7.10b). Three of these studies appeared to be updated versions of the same Cochrane review

and the fourth study was the IPD meta-analysis used for satisfaction with LNG-IUS (184).

The most recent Cochrane review (193) and the IPD meta-analysis (184) were evaluated

further. The IPD study was found to include 12 month satisfaction data from a larger overall

population of women. It also had the benefit of using IPD data, so was selected as the

preferable study. The reported dissatisfaction rates for second generation endometrial ablation

110/1034 were converted to satisfaction rates which were then used to calculate confidence

170

intervals (0.893, 95% CI 0.874-0.912). These values were used in the economic analysis for

satisfaction after endometrial ablation for DUB.

An electronic database search was performed (Appendix 7.10c-7.10d) to identify studies that

reported outcome after endometrial ablation in the presence of fibroids. Once duplicates had

been removed 315 studies were identified of which 17 were selected for further evaluation

because they reported the use of endometrial ablation in the presence of fibroids. None of the

studies reported satisfaction after second generation endometrial ablation in the presence of

intramural or subserosal fibroids, however one study reported that the presence of a small

fibroid uterus did not increase the hysterectomy rate (219). Once again this was extrapolated

and the same data were used for small fibroid uteri as for satisfaction after second generation

endometrial ablation for DUB. In the absence of published data for large fibroid uteri, the

gynaecologist panel were asked for their estimation of satisfaction rates, 1 year post

endometrial ablation. The median value of 0.575 was used as the point estimate and the range

of values as the data for sensitivity analysis (0.075-0.85)

Hysterectomy

The IPD meta-analysis for treatment satisfaction data following the use of a LNG-IUS and

after an endometrial ablation also reported data for hysterectomy so these values were used

for satisfaction after hysterectomy for DUB (184). The rates reported were 409/432 patients

satisfied which equates to a value of 0.95 and confidence intervals of 0.93 to 0.97. The studies

included in this IPD looked at heavy menstrual bleeding and included women with polyps and

fibroids. Only three studies supplied data regarding treatment outcome in the presence of

focal uterine pathology. The presence of endometrial polyps and fibroids were found not to be

statistically significant indicators of outcome. Thus it was assumed that satisfaction would be

171

the same whether hysterectomy was performed for DUB, polyps/SMFs or fibroids less than

12 weeks size.

For large fibroids (> 12 weeks uterine size), database searches were performed to identify

studies reporting satisfaction after hysterectomy for fibroids. No systematic reviews were

identified that examined this outcome directly. However a study of 397 women, that

retrospectively followed up women who had either had hysterectomy or uterine artery

embolisation reported that 88% of women who had undergone hysterectomy felt that their

symptoms were better and that 70% would recommend their treatment to a friend (220). A

disadvantage of this study was that the mean follow up time was 8.6 years. By scrutinising the

reference list of this study, two similar studies (195;221) were identified. The first was a

randomised trial of uterine artery embolisation (UAE) versus hysterectomy, with 51 women

in the hysterectomy arm (195). Women were included if they had fibroids of at least 2cm in

diameter (no upper limit), which caused symptoms and which a clinician thought justified

surgical treatment. At 12 months 93% of women would recommend their treatment to a

friend. The second study randomly allocated women with uterine fibroids up to 10cm in size

and menorrhagia, to two groups. The women in group 1 were offered UAE as an alternative to

hysterectomy for their fibroids and the women in group 2 were not offered the alternative and

all had hysterectomy (221). In total 17 women underwent hysterectomy and at 6 months 88%

reported that they would have the same treatment again, suggesting that they were satisfied

with the surgery (221). After evaluating the data from these studies 0.88 was chosen as the

satisfaction level as this was reported in the large retrospective study (220) and supported by

the smaller randomised study (221). For the sensitivity analysis, the proportion of women

172

who would recommend their treatment to a friend from the two separate studies was used

(0.70 and 0.93) (195;220).

No studies were identified that reported satisfaction rates after hysterectomy for endometrial

disease in premenopausal women. It was assumed that women having hysterectomy for

atypical hyperplasia or cancer would be 100% satisfied because they would be prevented from

developing cancer or treated for their cancer. For sensitivity analysis the lowest hysterectomy

satisfaction value (i.e. 0.88 for fibroids) and the highest satisfaction value (1.0 for cancer)

were used. For complex endometrial hyperplasia the same values as for DUB was used

because complex hyperplasia is a benign condition so can be grouped with other benign

causes of HMB without organic pathology i.e. DUB. Secondly, treatment satisfaction after

LNG-IUS was the same for complex hyperplasia as for DUB and so this extrapolation

regarding hysterectomy does not seem unreasonable. For endometrial disease overall, a

composite value of the atypia/cancer value and the hyperplasia value was calculated as 0.97.

For sensitivity analysis the lowest and highest satisfaction rates from the two categories were

used i.e. 0.95 for hyperplasia to 1.0 for cancer.

Polyp/SMF removal

Satisfaction after removal of endometrial polyps and submucous fibroids was calculated as a

composite of values for the two pathologies. For endometrial polyps two systematic reviews

were selected (166;167) from the 216 studies identified by database searches. Neither

systematic review included any meta-analysis because of the heterogeneity of the studies. The

more recent review (167) identified all of the studies used in the older one (166) as well as

more recently conducted studies. Satisfaction with polyp removal for abnormal uterine

bleeding was reported as 75-100% so he high quality studies were examined for the most

173

appropriate data. Three studies were prospectively conducted (117;222;223). The first study

was a cohort-controlled study comparing the effectiveness of outpatient and day case

endometrial polypectomy which included 58 women, predominantly postmenopausal. At 6

months, 34 women responded to a follow-up questionnaire that asked them about satisfaction

with the treatment. 78% of women in the outpatient group and 88% women from the inpatient

group were satisfied with their treatment, which equates to an 82% satisfaction rate overall

(117). The second trial was a cohort study which looked at 21 women with abnormal

menstrual bleeding and evaluated the change in their symptoms following endometrial

polypectomy (222). At 6 months there was a statistically significant reduction in menstrual

blood loss (p<0.001). Thirteen of the women had HMB, which persisted in 10 (77%) of them

at 6 months although they had a statistically significant reduction in pictorial blood loss

assessment chart (PBAC) scores (p=0.001). Overall, 86% of the women felt that they were

cured or that their symptoms had been relieved by endometrial polypectomy (222). The third

study randomised 150 premenopausal women with abnormal uterine bleeding to polypectomy

or conservative management and found no difference in PBAC scores at 6 months. There was

a significant difference between the groups for some of the secondary outcome measures

including mean periodic blood loss measured by visual analogue scale (p=0.02) and

occurrence of gynaecological symptoms (intermenstrual bleeding, pain), however, satisfaction

was not reported and so could not be used (223). As the two comparative studies reporting

satisfaction were small, the populations were combined and meta-analysed using Excel and

RevMan to calculate a satisfaction value of 0.86.

For satisfaction with transcervical resection of submucous fibroids (TCRF), 32 studies were

identified by medical database searches. Nine studies reported satisfaction after transcervical

174

myomectomy but seven were rejected because they had fewer than forty patients (5;224) or

reported satisfaction data from more than 3 years after treatment (170;225-228). The two

remaining studies (168;171) were both prospectively conducted and had populations of over

100 women, with more than 90% being premenopausal. The first study examined resection of

submucous fibroids using a resectoscope under general anaesthesia (90%) or local anaesthesia

with sedation. The average follow-up period was 2.3 years. Satisfaction with treatment was

reported as 71.4% (168). The second study examined removal of fibroids using a bipolar

intrauterine operating system and performed 38% of them under local anaesthesia. The

average follow-up period was 2.6 years. Satisfaction with treatment was reported as 86%

(171). In the decision tree resection as day-case surgery, under general anaesthesia was used

for submucosal fibroid removal as it is the gold standard. Therefore the satisfaction rate from

the first study was chosen as it appeared to be most appropriate for the analysis (168).

The data from the polypectomy meta-analysis and the selected submucous fibroid resection

study were used to create a value for the combined diagnostic group ‘polyps/SMF’ weighting

the data according to the disease prevalence within the group (50% polyps, 50% SMFs). The

value calculated was 0.79 and was used as the satisfaction value for removal of polyps and

submucosal fibroids.

To account for removal of erroneously diagnosed focal pathology (i.e. when no intrauterine

pathology was present so normal endometrium is being resected) a satisfaction rate for

dilatation and curettage was sought as the patients will essentially be having normal

endometrial tissue removed. 274 studies were identified form database searches but none

reported patient satisfaction following the procedure. One study was identified which

175

reported relief of HMB after D&C and stated that the menstrual blood loss was reduced for

the first month but returned to pre-operative levels after that (229). This does seem plausible

as D&C is primarily a diagnostic procedure and removing the superficial endometrium will

only be therapeutic until it grows back. As no additional data were available this value was

used and allocated a satisfaction score of zero for satisfaction following ‘virtual removal’

when no intrauterine lesion was present.

Myomectomy and uterine artery embolisation

The alternative analysis which looked at treating women who wished to preserve their fertility

included UAE and myomectomy as treatment options. Searches were performed to look at

patient satisfaction after both treatments at 1 year (Appendix 7.14-7.15). For myomectomy

one systematic review (230) was identified and selected from 120 studies and for UAE one

systematic review (231) was identified and selected from 169 studies. Both systematic

reviews compared UAE to surgical treatments (hysterectomy and myomectomy) (230;231)

and both found no difference in patient satisfaction or quality of life between UAE,

hysterectomy and myomectomy. Therefore the same value for satisfaction after hysterectomy

was used for UAE and myomectomy (0.88).

Costs

Cost values were mainly taken from Healthcare Resource Group (HRG) codes for 2009 -

2010. The ‘national average unit cost’ was used as the cost for each intervention. The

diagnostic and treatment codes for hysteroscopy include the cost of the consultation as well as

any diagnostic or therapeutic procedures but the other diagnostic test codes equated to just the

cost of the test. Within the decision tree patients could undergo multiple diagnostic tests and

176

treatments at one appointment. However, if the relevant HRG code costs were then used for

each aspect the cost of the consultation would be included multiple times. In order to give an

accurate reflection of the additional costs of multiple tests including hysteroscopic treatment,

cost of consultation was removed from the diagnostic hysteroscopy costs and also from the

therapeutic hysteroscopy codes. The cost of consultation and diagnostic hysteroscopy were

subtracted so that the value remaining was the additional cost of performing the therapeutic

hysteroscopic procedure alone. The relevant costs could then be added up depending upon the

tests and treatments that the patients had undergone. For example,

Cost of new gynaecology consultation £ 139

Cost of outpatient hysteroscopy (consultation+ diagnostic OPH) £ 216

Cost of outpatient hysteroscopic polypectomy £ 263

(consultation+ diagnostic OPH+ polypectomy)

Cost of transcervical resection of fibroid £1344

Cost of transvaginal ultrasound (test only) £ 55

Cost of GnRH analogues £ 226

So the cost of the diagnostic hysteroscopy is actually £216 - £139 = £ 77

And the additional cost of hysteroscopic polypectomy is £263 - £216 = £ 47

If a woman (assigned to the TVS diagnostic pathway) comes to clinic and has a diagnosis of

‘polyp/SMF’ made by TVS and then goes on to have removal of the lesion, the costs will

equate to all the women having a consultation and a scan, half of them having an outpatient

hysteroscopic polypectomy (ratio of polyps to submucous fibroids 50:50) and half of them

177

returning at a later date for a scheduled transcervical resection of fibroid (with 70% of these

women having endometrial pre-treatment with GnRH analogues.

cGynaeNew + cTVS + (0.5*cOPH) + (0.5*cPolypectomy) + (0.5*cTCRF) + (0.35*cGnRH)

= £139 + £55 + (0.5 x £77) + (0.5 x £47) + (0.5 x £1344) + (0.35 x £226)

= £139 + £55 + £38.50 + £23.50 + £672 + £79.10

= £1007.10

However, if the HRG code was used for hysteroscopic polypectomy alone the cost would be

considered as £263 or if the code for TCRF was used it would be £1344. For the cost of LNG-

IUS and GnRH-a costs were taken from the British National Formulary (BNF) (25) The cost

of uterine artery embolisation could not be identified in the HRG codes so the value used in

the decision tree came from the REST study (195) as reported and used by NICE in an

economic analysis for treatment of fibroids in their HMB guideline (142). Although this cost

was published in 2007, the cost stated for hysterectomy was comparable to the cost stated in

the HRG codes 2009-2010 (£2,566 vs. £2961) which are used in this economic analysis

therefore no adjustment was made for inflation. For the cost of a GP appointment, data were

taken from the Personal Social Services Research Unit (PSSRU (232)). Table 8.12 details all

of the costs, ranges and sources of data.

178

Table 8.12 Costs for investigation and treatment of women presenting with HMB Variable Decision tree name Cost (£) Lower

quartile

Upper

quartile

Explanation

Consultations

GP visit cGP_visit 36 Personal Social Services Research Unit. (PSSRU)

http://www.pssru.ac.uk/uc/uc2010contents.htm 11min

appt. Inc direct staff

New gynaecology

outpatient appointment

cGynaeNew 139 111 161 502 HRG consultant service code

Follow-up gynaecology

outpatient appointment

cGynaeFU 97 73 109 502 HRG consultant service code

Diagnostic tests

Diagnostic outpatient

hysteroscopy

cOPH 77 43 111 MA10Z HRG less cost of new gynaecology outpatient

appointment


cTVS 55 40 66 HRG code RA23Z

Endometrial biopsy cEBx 34 16 34 HRG code DAP824

Saline infusion scan

cSIS 71 54 83 HRG code RA24Z

Confirmatory test cConfirmatory_test 115.72 77.56 130.66 (0.91*cEBx)+(0.09*cDandC)

Treatments

LNG-IUS fitting

cLngIUS_fitted 85.66

Endometrial polypectomy

cPolypectomy 47 0 120 MA12Z HRG less the cost of diagnostic hysteroscopy

MA10Z.

Transcervical resection of

fibroid with D+C cost

cTCRF 1344 981 1541 HRG code MA09Z

179

Table 8.12continued

Variable Decision tree name Cost (£) Lower

quartile

Upper

quartile

Explanation

Transcervical resection of

fibroid without D+C cost

cTCRF_noDandC 401 281 433 HRG code MA09Z less the cost of D&C MA10Z

Use of GnRH-a analogue

cGnRHanalogue 225.73 BNF(25)

Hysterectomy for benign

disease

cHysterectomy 2961 2346 3406 HRG code MA07D

Hysterectomy for malignant

disease

cHysterectomy_malignant 3898 3052 4464 HRG code MA06Z

Second laparotomy after

cancer diagnosis to remove

ovaries, lymph nodes etc.

cReturn_for_BSO 3898 3052 4464 HRG code MA06Z


cEndoAblation 896 697 1024 HRG code MA12Z

Dilatation and curettage

cDandC 942 700 1108 HRG code MA10Z

Rescue treatment

cRescue_Treatment 3136 2493 3606 cGP_visit + cGynaeNew + cHysterectomy

Myomectomy

cMyomectomy 2961 2346 3406 HRG code MA07D

UAE cUAE 1685 1465 1905 REST study (195) MA09Z Upper Genital Tract Laparoscopic / Endoscopic Intermediate Procedures MA10Z Upper Genital Tract Laparoscopic / Endoscopic Minor Procedures

MA11Z Upper Genital Tract Intermediate Procedures MA12Z Resection and ablation procedures for intra-uterine lesions

MA06Z Open Major Upper and Lower Genital Tract Procedures with malignancy MA07D Upper Genital Tract Major Procedures without Major CC

RA 23Z Ultrasound Scan less than 20 minutes RA 24Z Ultrasound Scan more than 20 minutes

DAP824 Histology / Histopathology 502 Face to face, non-admission gynaecology consultant appointment

OPH= outpatient hysteroscopy; TVS = transvaginal scan; EBx= endometrial biopsy; SIS= saline infusion sonography; SMF= submucous fibroid; DUB= dysfunctional uterine bleeding;

LNG-IUS= levonorgestrel intrauterine system; TCRF= transcervical resection of fibroid; D&C= dilatation and curettage; BSO= bilateral salpingo-oophorectomy; UAE= uterine artery

embolisation; GnRH-a= Gonadatrophin releasing hormone analogue; BNF= British National Formulary; GP= general practitioner; SR= systematic review; CI= confidence interval;

HRG= health resource group.

180

Methods for the cost-effectiveness analysis

(The following explanation of the methods for the economic analysis was written with Dr P. Barton, Reader

in Health Economics. This section is essential for understanding the figures displayed in the results section

and therefore is included within this thesis.)

A decision tree was constructed using TreeAge Pro 2009. Model inputs included probabilities

and costs. Effectiveness was measured in terms of patient satisfaction. Therefore each branch

of the tree had an effectiveness outcome of 1 for a positive outcome and 0 if otherwise. Costs

were unit costs for the various tests and treatments and were therefore treated as known with

certainty, while probabilities depended on data and were treated as uncertain, to be varied in

probabilistic sensitivity analysis. Note that the uncertainty in overall effectiveness for a given

strategy is fully accounted for in the uncertainty in the probabilities, so there is no need to

vary the outcome parameters.

The model was first run using the point estimates of the branch probabilities. The results,

known as “base case” results, are shown in terms of mean costs and effectiveness (overall

proportion of positive outcomes) for each strategy modelled. These are tabulated and shown

in a cost-effectiveness plane, with the mean cost shown on the vertical axis and the mean

effectiveness shown on the horizontal axis. In some cases, a further plot was made of selected

strategies to show more clearly the relationship between points that were close together on the

first graph.

Any strategy which has greater cost and worse effectiveness than some other strategy is said

to be simply dominated. Such a strategy can be excluded from consideration. An incremental

cost-effectiveness ratio (ICER) can be calculated between any two non-dominated strategies.

The ICER is calculated by dividing the difference in cost by the difference in effectiveness. If

181

the ICER is less than the maximum willingness to pay for an additional positive outcome,

then the more effective strategy can be said to be cost-effective relative to the other strategy.

There is a further reason for excluding strategies known as extended dominance. This can

only apply when there are three or more non-dominated strategies. In this case, two different

strategies (incorporating a cheaper but less effective strategy and a more effective, more

costly strategy relative to a third strategy) can be mixed together, with a proportion of patients

getting one or other of the strategies such that the third strategy now becomes dominated (i.e.

is more expensive and less effective) than the blended strategies Suppose that A, B and C are

non-dominated strategies in order of increasing cost. Since they are non-dominated, they must

also be in order of increasing effectiveness. Now suppose that the ICER of B over A is higher

than the ICER of C over A. Then, if B is cost-effective compared to A, so also must C be

cost-effective compared to A and B. In such a case, there is no value of willingness to pay per

positive outcome at which B will be the preferred strategy, and the strategy B can be excluded

by extended dominance. Extended dominance can be seen on a cost-effectiveness plane. The

point for strategy B will be above the straight line joining the points for strategies A and C.

Once all dominated strategies have been excluded, whether for simple or extended

dominance, the remaining strategies are potentially cost-effective. Which will be preferred

depends on the willingness to pay for an additional positive outcome.

To test for the effect of uncertainty in the model inputs, two types of sensitivity analysis were

used: probabilistic and deterministic.

182

In probabilistic sensitivity analysis, probability distributions are placed around the point

estimate for each model parameter. If there is correlation between the uncertainties, joint or

conditional distributions may be used. Beta distributions were used to represent the

uncertainty around the branch probabilities in this model. The beta distribution is the standard

distribution for a proportion. It has two parameters a and b, with mean a/(a+b) and variance

essentially decreasing as a and b increase.

For individual parameters of the models, the information available was in the form of a point

estimate and a 95% confidence interval. In all cases, the beta distribution was selected with

mean equal to the point estimate. Usually the distribution also matched the width of the

confidence interval, but there were two main exceptions to this:

If the point estimate is either 0 or 1, then it is not possible to find a beta distribution

with that mean. While it can be argued that the true mean estimate of the probability

should be strictly between 0 and 1, taking any actual numerical value would risk

overcompensating. Accordingly, it was decided to treat such probabilities as fixed,

thereby preserving the mean but slightly underestimating the uncertainty in the model.

If either parameter a or b is less than 2, the beta distribution gives an unreasonably

high proportion of values very close to the extreme values 0 or 1. To avoid this, in

such cases the values of a and b were increased to preserve the mean of the

distribution but ensure that both values were at least 2. Again, this slightly

underestimates the overall uncertainty.

In some cases, only a point estimate was available. In such cases, it is not appropriate to

assume that the value is fixed. Instead the widest possible uncertainty was modelled subject to

183

the constraint that both parameters a and b should be at least 2, for the reasons given in the

previous paragraph.

In the analysis of SIS for detecting fibroids, point estimates of TPR and FPR were assumed to

be the same as for TVS. In these cases, independent samples were taken from Beta

distributions with the same parameters a and b.

Since the costs in the model are all unit costs of specific procedures, these were treated as

fixed, and the only parameters to be varied were the probabilities in the tree, which are

proportions of patients expected to follow each branch.

When the model was run for probabilistic sensitivity analysis (PSA), 1000 replications were

made, sampling from distributions for all branch probabilities simultaneously. It is generally

accepted that 1000 replications are sufficient to give a clear picture of the uncertainty.

The parameters for this Beta distribution used for the probabilistic sensitivity analysis of the

HMB tree are shown in Table 8.13.

184

Table 8.13 Beta distribution parameters for analysis of the base case decision tree for

investigating women with heavy menstrual bleeding Parameter a* b* Low High

Prevalence of disease

EndometrialDisease 5 95 0.02 0.10

FibroidLarge 6 94 0.02 0.11

FibroidSmall 19 81 0.12 0.27

PolyporSMF 38 62 0.29 0.48

Success rates

EBx 728 72 0.89 0.93

OPH 67.9 2.1 0.92 1.00

SIS 1520 80 0.94 0.96

TVS 198 2 0.97 1.00

True Positive Rates

ED_EBx 28.86 8.14 0.64 0.90

ED_OPH 1248 352 0.76 0.80

ED_SIS 2.03 4.97 0.05 0.65

ED_TVS 3.249 2.451 0.19 0.90

Fibroids_SIS 38 2 0.87 0.99

Fibroids_TVS 38 2 0.87 0.99

Normal_EBx 95 5 0.90 0.98

Normal_OPH 1424 176 0.87 0.90

Normal_SIS 264 36 0.84 0.91

Normal_TVS 258 42 0.82 0.90

PolyporSMF_EBx 2.16 5.84 0.04 0.60

PolypsorSMF_OPH 227.5 22.5 0.87 0.94

PolypsorSMF_SIS 104.4 15.6 0.80 0.92

PolypsorSMF_TVS 22.5 27.5 0.32 0.59

False Positive Rates

CompHyp_EBx 7.5 142.5 0.02 0.09

ED_OPH 1000 24000 0.04 0.04

ED_SIS 2 98 0.00 0.05

ED_TVS 27.2 52.8 0.24 0.45

Fibroids_SIS 2.16 9.84 0.03 0.43

Fibroids_TVS 2.16 9.84 0.03 0.43

PolypSMF_EBx 2 998 0.00 0.01

PolypSMF_OPH 33 517 0.04 0.08

PolypSMF_SIS 3.9 26.1 0.04 0.27

PolypSMF_TVS 7.7 27.3 0.10 0.37

pExDetectsFibroids 8 2 0.52 0.97

pHyperplasia 3 2 0.19 0.93

185

Table 8.13 continued Parameter a* b* Low High

Probability of being satisfied

EA_DUB 924 110 0.87 0.91

EA_Fibroids 2.76 2.04 0.17 0.92

Hysterectomy_AtypCa parameter fixed at value 1

Hysterectomy_ED 164.9 5.1 0.94 0.99

Hysterectomy_Fibroids 24.64 3.36 0.74 0.97

Hysterectomy_HMB 380 20 0.93 0.97

Hysterectomy_Hyperpl 380 20 0.93 0.97

LngIUS_DUB 107.9 22.1 0.76 0.89

LngIUS_ED 9.68 12.32 0.24 0.65

LngIUS_Fibroids 2.03 4.97 0.05 0.65

LngIUS_Hyperplasia 91.3 18.7 0.75 0.89

LngIUS_PolypSMF 420 580 0.39 0.45

Removal 11 3 0.55 0.95

Removal_CavityNormal parameter fixed at value 0

a, b: Parameters of Beta distribution. Low, High: 2.5 and 97.5 percentiles respectively of fitted distribution, corresponding to

lower and higher limits of 95% confidence interval. All parameters sampled independently except for the prevalence

parameters, which were sampled together from a Dirichlet distribution (see text). OPH= outpatient hysteroscopy; TVS =

transvaginal scan; EBx= endometrial biopsy; SIS= saline infusion sonography; SMF= submucous fibroid; DUB=

dysfunctional uterine bleeding

*Please read ‘Methods for the cost-effectiveness analysis', for explanation regarding the parameters a and b

For models (such as Markov models) in which there is a non-linear relationship between

model inputs and outputs it is appropriate to give a table of mean results from the probabilistic

sensitivity analysis, as the Bayesian viewpoint is that the mean results from the PSA are the

appropriate basis for decision making. However, in the case of the models presented here,

these results would be statistically equivalent to the base case results and therefore there is no

need to produce such a table. Results that have been shown are as follows:

A cost-effectiveness scattergraph: This shows, on a single graph, the uncertainty in the

absolute expected cost and effectiveness for each option separately. For each option, the

results of the 1000 replications of the model were shown each as a single point. In practice,

the printed symbols used merge to form a “cloud” giving the general range of uncertainty in

the results. The vertical spread of this cloud reflects the uncertainty in the overall cost and the

186

horizontal spread the uncertainty in overall effectiveness, while the centre of the cloud, where

the points are most densely packed, indicates the most likely cost and effectiveness.

A cost-effectiveness acceptability frontier (CEAF): At any given willingness to pay, the

preferred option is determined by the mean outcomes, which in the case of the models here

are the same as the base case results described earlier. The CEAF shows the proportion of

model replications for which this option remained the preferred strategy, as a function of the

willingness to pay for an additional positive outcome.

While the graphs described above are the only convenient ways of showing results comparing

all the modelled options, they are applicable only to a decision in which exactly those options

are included. For other purposes, it is helpful to look at pairwise comparisons between

strategies. The results shown from a pairwise comparison are helpful to any decision problem

in which both those strategies are included. Pairwise comparisons were made between

successive non-dominated options (in order of increasing cost or effectiveness), and others

where a dominated option was close to another option.

For pairwise comparisons, the incremental cost-effectiveness scattergraph was shown. In this

type of graph, there is a single point for each of the 1000 replications of the model, showing

the difference in cost and effectiveness between the two strategies. If the “clouds” shown in

the cost-effectiveness scattergraph for the two strategies overlap, it may be for two possible

reasons. First, it may be because there is genuine uncertainty as to which is the more costly

and/or more effective strategy. Second, it may be that one strategy is consistently more costly

and/or more effective than the other, but this consistent difference is small compared to the

187

uncertainty in the absolute costs and/or effectiveness. The incremental cost-effectiveness

scattergraph distinguishes between these two cases and shows the relevant uncertainty for a

decision maker.

The other graph plotted for the pairwise comparisons was the cost-effectiveness acceptability

curve (CEAC). This shows the proportion of model replications in which one of the strategies

is cost-effective compared to the other, across a range of values of willingness to pay per

additional positive outcome.

In deterministic sensitivity analysis, one or more model inputs are varied systematically and

the effect on the model outcomes is noted. This was used to test the effect of reducing the

prevalence of polyps and submucous fibroids and increasing the prevalence of dysfunctional

uterine bleeding and also to examine the effect when the unit cost of saline infusion

sonography was reduced. For prevalence of the various pathologies, a Dirichlet distribution

was used. This is the generalisation of the Beta distribution for more than two options. Given

that the prevalence data came from different sources, it was necessary to take a compromise

between the effective sample sizes indicated by those sources. An effective sample size of 100

was assumed. The distribution of any prevalence parameter on its own then follows a Beta

distribution

188

CHAPTER 9

RESULTS OF THE ECONOMIC ANALYSIS OF DIAGNOSTIC

STRATEGIES FOR THE INVESTIGATION OF HEAVY

MENSTRUAL BLEEDING

Deterministic results: Base case

The current recommended first line treatment for HMB in women not desiring immediate

fertility, is the levonorgestrel intrauterine system treatment (LNG-IUS) otherwise known as

the Mirena® coil (Bayer HealthCare Pharmaceuticals Inc. Wayne, USA) (142). The majority

of women with HMB have no uterine pathology (known as dysfunctional uterine bleeding) or

benign uterine pathologies such as small uterine fibroids or endometrial hyperplasia and all

these conditions respond well in general to the LNG-IUS (161;162).

Thus, given the negligible chance of life-threatening disease, lack of recommendations

stipulating the need for routine diagnostic testing and the known effectiveness and

applicability of the LNG-IUS in HMB, the strategy using LNG-IUS without any preliminary

diagnostic testing was chosen as the reference strategy. The costs and effects of the other

diagnostic testing strategies were compared against this reference strategy. In addition, a

recent cost-effectiveness analysis informed by IPD analysis of published trials for treatment

of HMB suggested that surgical treatment with hysterectomy was more cost-effective than the

189

LNG-IUS(136) so a strategy of hysterectomy without diagnostic work up was also

considered.

Table 9.1 reports the deterministic results, referencing all other diagnostic or treatment

strategies to the LNG-IUS treatment alone base-case strategy.

Table 9.1 Determinist analysis results for the investigation of women with heavy

menstrual bleeding

Strategy Cost (£) Effectiveness

(satisfaction)

LNG-IUS alone 1067 0.9333

OPH alone 1078 0.9641

SIS alone 1083 0.9629

TVS alone 1085 0.9551

TVS + OPH 1139 0.9644

OPH and Ebx 1149 0.9674

SIS + OPH 1170 0.9645

EBx alone 1209 0.9460

SIS + Ebx 1223 0.9643

TVS+ OPH+Ebx 1227 0.9649

TVS + Ebx 1231 0.9539

SIS +OPH +Ebx 1256 0.9650

Hysterectomy alone 3182 0.9335 EBx = endometrial biopsy; OPH = outpatient hysteroscopy; SIS = saline infusion sonography; TVS = transvaginal ultrasound

scan; LNG-IUS = levonorgestrel intrauterine system.

Outcomes

Direct treatment without preliminary diagnostic testing was less effective than treatment

instigated after diagnostic testing. The least effective approach was the base case strategy of

LNG-IUS treatment alone followed closely by surgical treatment with hysterectomy, both

approaches resulting in satisfaction rates around 93.3-93.4%. The effectiveness of HMB

management was similar across all testing strategies ranging from 94.6% to 96.7% rates of

190

satisfaction. The most effective strategy was combination testing with outpatient hysteroscopy

and endometrial biopsy.

Costs

The LNG-IUS treatment alone base case strategy was the cheapest costing £1067 per woman

treated for HMB in a secondary care setting and the strategy of hysterectomy for all women

was the most expensive at £3182 per woman treated i.e. £2116 more than the approach of

LNG-IUS treatment alone. The cheapest of the nine diagnostic testing strategies was the use

of outpatient hysteroscopy alone, costing £1078 for every woman treated, i.e. £11 more than

universal LNG-IUS treatment.

Cost-effectiveness and dominance

Only the testing strategies outpatient hysteroscopy (OPH) alone and outpatient hysteroscopy

combined with endometrial biopsy (OPH + EBx) remain non-dominated by alternative

empirical treatment or diagnostic testing strategies. It is clear from the analysis that the

strategy OPH alone dominates the testing strategies SIS alone and TVS alone. The

combination testing strategy TVS and OPH is excluded by extended dominance between OPH

alone and OPH + EBx. The remaining seven alternate strategies are dominated by OPH +

EBx. Table 9.2 presents the deterministic analysis restricted to the non-dominated competing

strategies.

191

Table 9.2 Deterministic results for the non-dominated strategies for investigation of

women with heavy menstrual bleeding

Strategy

Total

Cost

(£)

Incremental

Cost (£)

Effectiveness

(satisfaction)

Incremental

effectiveness

ICER

LNG-IUS

alone 1067

0.9333

OPH alone 1078 11 0.9641 0.0308 359

OPH and

EBx 1149 71 0.9674 0.0033 21500 Apparent anomalies with subtraction are due to rounding effects

EBx = endometrial biopsy; ICER = incremental cost-effectiveness ratio; LNG-IUS = levonorgestrel intrauterine system;

OPH= outpatient hysteroscopy

The cheapest strategy is the base-case scenario of no testing and universal treatment with the

LNG-IUS alone. The most effective strategy is the combination of initial testing with OPH +

EBx, but this comes at a greater cost, generating an ICER of £21,500 i.e. the strategy requires

an investment of £21,500 to gain an extra woman satisfied following treatment for HMB

compared with investigation with OPH alone. The single test strategy of OPH is slightly less

effective than the strategy of OPH with the addition of EBx, but is substantially less costly.

The ICER for OPH is £359, i.e. an additional financial outlay of £360 is necessary to acquire

an extra woman satisfied following treatment for HMB.

Figure 9.1 shows the total costs and effectiveness of alternative strategies for the diagnosis

and treatment of HMB in secondary care. The line presented graphically joins the non-

dominated strategies (OPH alone and OPH+EBx). Any strategy plotted above this line is not

considered cost-effective in relation to the non-dominated strategies.

192

Figure 9.1 Cost-effectiveness plane showing the results of deterministic analysis of the

strategies for investigation of women with heavy menstrual bleeding

LNG-IUS = levonorgestrel intrauterine system; EBx = endometrial biopsy; OPH = outpatient hysteroscopy; SIS = saline

infusion sonography; TVS= transvaginal ultrasound scan

It is clear that direct treatment with hysterectomy is the least cost-effective strategy.

Replicating the figure and excluding the hysterectomy strategy reduces the scale of the y axis

allowing closer examination of the remaining testing strategies.

Figure 9.2 reveals that the options "TVS alone" and "SIS alone" are sufficiently close to the

boundary of dominance that it is worth checking for the uncertainty between these dominated

alternatives in addition to the non-dominated options “OPH” and “OPH + EBx” through

sensitivity analyses.

0

500

1000

1500

2000

2500

3000

3500

0.93 0.94 0.95 0.96 0.97

Co

st (

£)

Effectiveness (satisfaction)

Deterministic Analysis (all strategies)

Mirena alone

OPH alone

OPH and EBx

TVS + OPH

SIS alone

TVS alone

SIS + OPH

EBx alone

SIS + EBx

TVS+ OPH+EBx

TVS + Ebx

SIS +OPH +EBx

Hysterectomy alone

Not Dominated

LNG-IUS alone

193

Figure 9.2 Cost-effectiveness plane showing the results of deterministic analysis of the

strategies for investigation of women with heavy menstrual bleeding- hysterectomy

excluded

LNG-IUS = levonorgestrel intrauterine system; EBx = endometrial biopsy; OPH = outpatient hysteroscopy; TVS=

transvaginal scan; SIS= saline infusion sonography

1050

1100

1150

1200

1250

1300

0.93 0.94 0.95 0.96 0.97

Co

st (

£)


Deterministic Analysis (Hysterectomy alone not shown)

Mirena alone

OPH alone

OPH and EBx

TVS + OPH

SIS alone

TVS alone

SIS + OPH

EBx alone

SIS + EBx

TVS+ OPH+EBx

TVS + Ebx

SIS +OPH +EBx

Not Dominated

LNG-IUS alone

194

Probabilistic sensitivity analysis results: Base case

Figure 9.3 demonstrates the uncertainty surrounding the absolute expected cost and

effectiveness of each of the strategies, with the true value lying somewhere within the ‘cloud’

of plotted points, probably where they are most densely clustered. It shows that hysterectomy

is too expensive to be a competitive option but that there is overlap between the remaining

strategies.

Figure 9.3 Scatterplot showing the uncertainty in costs and effectiveness within the

model for each of the individual strategies for investigation of women with heavy

menstrual bleeding

LNG-IUS = levonorgestrel intrauterine system; EBx = endometrial biopsy; OPH = outpatient hysteroscopy; TVS=

transvaginal scan; SIS= saline infusion sonography

Figure 9.4 illustrates the overall uncertainty related to the optimal decision across a range of

plausible willingness to pay (WTP) values, where the willingness to pay is measured in £ per

additional case satisfied. When a model is linear in all parameters and they are sampled

independently, the mean of the probabilistic results will be the same as the deterministic

0

500

1000

1500

2000

2500

3000

3500

4000

0.8 0.85 0.9 0.95 1

Co

st (

£)


CE Scatterplot

C(Mirena alone)

C(Hysterectomy alone)

C(OPH alone)

C(TVS alone)

C(EBx alone)

C(SIS alone)

C(OPH and EBx)

C(SIS + EBx)

C(SIS + OPH)

C(TVS + Ebx)

C(TVS + OPH)

C(SIS +OPH +EBx)

C(TVS+ OPH+EBx)

C (LNG-IUS alone)

195

result. This is true for the current model, as the only parameters which are not independently

sampled are the prevalences, and these do not interact in the calculation of the model results.

The cost-effectiveness acceptability frontier (CEAF) (Figure 9.4) is generated as follows.

First, for any willingness to pay, the optimal option is determined based on the mean results.

Then the proportion of model replications for which that was the optimal option is found and

plotted. For example, considering a WTP of £10,000 per case satisfied, the preferred option

based on the mean results is OPH alone, and this was optimal in around 60% of the model

replications. Thus there is an estimated probability of 40% that there is a better option than

OPH at that WTP. By definition, only the options which are not dominated in the mean results

can appear on the CEAF. Sometimes the probability shown will be lower than 50%. It will

often be the case that the option preferred on mean values is also the preferred option in the

highest proportion of model replications, but this not always so. As the willingness to pay

crosses the ICER between two non-dominated options, the choice of optimal option changes,

and there will usually be a discontinuity in the curve.

Figure 9.4 Cost-effectiveness acceptability frontier showing the optimal investigative

strategy across a range of willingness to pay thresholds for the sensitivity analysis of

stratgies to investigate women with heavy menstrual bleeding

LNG-IUS = levonorgestrel intrauterine system; EBx = endometrial biopsy; OPH = outpatient hysteroscopy

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P C

ost

-Eff

ect

ive

(O

pti

mal

Str

ate

gy O

nly

)

Willingness to Pay (£ / additional case satisfied)

LNG-IUS alone

OPH alone

OPH and EBx

196

For the current model, the CEAF shows the same range of preferred options already shown in

Table 9.2, but indicates that there is appreciable uncertainty about the preferred option across

the whole range of WTP values plotted. To explore the uncertainty more fully, it is helpful to

consider a range of appropriate pair wise comparisons between the different options.

Comparisons are shown between adjacent non-dominated options. There are also options that

are dominated on mean values, but whose mean values are close to the non-dominance lines

shown on Figure 9.2. These options are compared to relevant non-dominated options.

197

Outpatient hysteroscopy versus levonorgestrel releasing intrauterine system

The cost-effectiveness plane (Figure 9.5a) shows the modelled uncertainty in the difference in

costs between OPH alone and LNG-IUS alone. It shows that OPH alone is consistently more

effective than LNG-IUS alone, and is likely (but not certain) to be more costly. The cost-

effectiveness acceptability curve (CEAC) (Figure 9.5b) shows the proportion of model

replications for which OPH alone is preferred to LNG-IUS alone at any given WTP. OPH is

the preferred option at any willingness to pay over £360 per additional case satisfied, but there

is considerable uncertainty when the WTP is just above this figure. However, by the time the

WTP exceeds £8,000 per additional case satisfied, it is almost certain that OPH is preferred to

LNG-IUS.

Figure 9.5 Cost-effectiveness plane (a) and cost-effectiveness acceptability curve (b):

Outpatient hysteroscopy alone strategy relative to the LNG-IUS alone strategy

-250

-200

-150

-100

-50

0

50

100

150

200

250

0 0.05 0.1 0.15

Incr

em

en

tal C

ost

(£

)

Incremental Effectiveness

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P C

ost

-Eff

ect

ive

Willingness to Pay (£/additional case satisfied)

(a) (b)

198

Outpatient hysteroscopy and endometrial biopsy versus outpatient hysteroscopy

The graph (Figure 9.6a) shows the modelled uncertainty in the difference in costs between

"OPH and EBx" and OPH alone. It shows that adding EBx to OPH increases the cost and is

very likely to increase the effectiveness. The CEAC (Figure 9.6b) shows the proportion of

model replications for which OPH and EB is preferred to OPH alone at any given willingness

to pay per additional case satisfied. It is more likely than not that OPH + EBx is cost-effective

compared to OPH above a WTP threshold of around £23,000. However, there is considerable

uncertainty throughout the range of WTP values shown. About 30% of replications favour

OPH alone even if the WTP is as high as £40,000 per additional case satisfied.


Outpatient hysteroscopy and endometrial biopsy strategy relative to the outpatient

hysteroscopy alone strategy

0

20

40

60

80

100

120

-0.01 0 0.01 0.02

Incr

em

en

tal C

ost

(£

)


00.10.20.30.40.50.60.70.80.9

1

0 10000 20000 30000 40000

P O

PH

an

d E

Bx

Co

st-E

ffe

ctiv

e


(a) (b)

199

Assessment of potentially cost-effective competing strategies

In addition to the non-dominated testing strategies “OPH” and “OPH + EBx”, the single

testing strategies TVS and SIS were sufficiently close to the boundary of dominance to

explore the level of uncertainty within the model pertaining to these two dominated strategies.

Transvaginal ultrasound scan versus levonorgestrel releasing intrauterine system


TVS alone and LNG-IUS alone. It shows that TVS alone is consistently more effective than

LNG-IUS alone, and is likely (but not certain) to be more costly. The CEAC (Figure 9.7b)

shows that it is more likely than not that TVS is cost-effective compared to LNG-IUS above a

WTP threshold of around £1000 and this is almost certain at WTP thresholds beyond £9000.


Transvaginal ultrasound scan alone strategy relative to the LNG-IUS alone strategy

-250

-200

-150

-100

-50

0

50

100

150

200

0 0.05 0.1

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 20000 40000

P C

ost

-Eff

ect

ive


(a) (b)

200

Saline infusion sonography versus levonorgestrel releasing intrauterine system


SIS alone and LNG-IUS alone. It shows that SIS alone is consistently more effective than

LNG-IUS alone, and is likely (but not certain) to be more costly. The CEAC (Figure 9.8b)

shows that the likelihood is that SIS is cost-effective compared to LNG-IUS above a WTP

threshold of around £800. This is almost certainly the case at WTP thresholds beyond £8000.


Saline infusion sonography alone strategy relative to the LNG-IUS alone strategy

-200

-150

-100

-50

0

50

100

150

200

0 0.05 0.1

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P C

ost

-Eff

ect

ive


(a) (b)

201

Outpatient hysteroscopy versus transvaginal ultrasound scan


OPH alone and TVS alone. It shows that OPH alone is almost certainly more effective than

TVS alone but it is unclear whether it is more costly. The CEAC (Figure 9.9b) shows the

proportion of model replications for which OPH alone is preferred to TVS alone at any given

willingness to pay per additional case satisfied. The likelihood is that OPH is cost-effective

compared to TVS above any WTP threshold and this is almost certain at WTP thresholds

beyond £9000.


Outpatient hysteroscopy alone strategy relative to the transvaginal ultrasound scan

alone strategy

-120

-100

-80

-60

-40

-20

0

20

40

60

80

-0.02 0 0.02 0.04

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P C

ost

-Eff

ect

ive


(a) (b)

202

Outpatient hysteroscopy versus saline infusion sonography


OPH alone and SIS alone. It shows that OPH alone is likely to be more effective than SIS

alone, and there is considerable uncertainty as to which is more costly. The CEAC (Figure

9.10b) shows the proportion of model replications for which OPH alone is preferred to SIS

alone at any given willingness to pay per additional case satisfied. The likelihood is that OPH

is cost-effective compared to SIS above any WTP threshold although there is considerable

uncertainty throughout. Even at a WTP of £40,000 per additional case satisfied, SIS is

preferred in 20% of model replications.


Outpatient hysteroscopy alone strategy relative to the saline infusion sonography alone

strategy

-120-100

-80-60-40-20

020406080

100

-0.01 0 0.01 0.02

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P C

ost

-Eff

ect

ive


(a) (b)

203

Deterministic sensitivity analysis results

Assessment of the impact of reducing the cost of saline infusion sonography

The probabilistic sensitivity analyses show uncertainty around whether OPH or SIS is the

most cost-effective investigative strategy. To assess this uncertainty further, deterministic

sensitivity analysis was performed to reduce the cost of SIS and determine at what cost it

would become more cost-effective than OPH. Table 9.3 details the ICER values when the cost

of SIS is reduced.

Table 9.3 Incremental cost-effectiveness ratio value for saline infusion sonography when

the cost is varied

SIS cost £

ICER

65 76

60 4006

55 7937

53 9509

52 10295

50 11867

45 15797

35 23658

SIS= saline infusion scan; ICER= incremental cost-effectiveness ratio

In this analysis, the unit cost for SIS was reduced from £71 (base case cost), keeping all other

variables fixed. When the cost was reduced to £65, the strategy "SIS alone" was no longer

dominated by "OPH alone". However, the modelled ICER was £76 per additional case

satisfied, suggesting that OPH alone is still highly cost-effective compared to SIS alone. As

the cost of SIS reduces further, the ICER increases (see Figure 9.11). Considering an

illustrative willingness to pay of £10,000 per case satisfied, the ICER goes above this figure

when the cost of SIS drops to £52. In that case, OPH is no longer cost-effective compared to

204

SIS, and SIS becomes the preferred strategy (of the two) on cost-effectiveness grounds. It

should also be noted that at a unit cost for SIS of £53 or lower, the strategy "SIS alone"

becomes less costly, as well as remaining more effective, than LNG-IUS alone.

Figure 9.11 Incremental cost-effectiveness ratios between outpatient hysteroscopy and

saline infusion sonography when the cost of saline infusion sonography is varied

SIS= saline infusion scan; ICER= incremental cost-effectiveness ratio

Assessment of the impact of the prevalence of focal uterine pathology

A high prevalence of intracavity focal endometrial lesions will favour outpatient hysteroscopy

over the imaging technologies TVS and SIS because it is more likely to diagnose the lesions

and treatment can be initiated with only a small additional cost during the diagnostic

procedure (the so called ‘see & treat’ approach). If the prevalence of endometrial polyps and

submucous fibroids is over-estimated within the decision tree, OPH will falsely appear the

most cost-effective. Similarly the prevalence of DUB may be an underestimate. To assess the

effect of prevalence on the analysis the prevalence of polyps / SMF’s was varied, keeping the

prevalence of fibroids and endometrial disease fixed. The prevalence of dysfunctional uterine

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

0 10 20 30 40 50 60 70

ICER

(£

/ a

dd

itio

nal

cas

e s

atis

fie

d)

Cost of SIS (£)

205

bleeding was changed inversely to compensate for the change in prevalence of polyp or SMF.

All other variables in the model were fixed at their point estimates. Figure 9.12 shows the

preferred option at a range of values of willingness to pay per additional case satisfied,

varying the prevalence of polyp or SMF. For example, at a prevalence of 30%, the

combination of OPH and EBx is preferred if the willingness to pay per additional case

satisfied is more than about £27,000, while OPH alone is preferred if this willingness to pay is

between £2000 and £27,000. Only at a willingness to pay below £2000 per case satisfied is

LNG-IUS alone preferred. For this prevalence, other options are dominated and so not

preferred at any willingness to pay value. It is only at a prevalence of polyps/SMFs of 27%

that SIS becomes a non-dominated option and at 24% when TVS becomes non-dominated.

Figure 9.12 Incremental cost-effectiveness ratios between non-dominated options at a

variety of prevalence values for polyps / submucous fibroids and dysfunctional uterine

bleeding

OPH = outpatient hysteroscopy; SIS = saline infusion sonography; TVS = transvaginal ultrasound scan; LNG-IUS =

levonorgestrel intrauterine system, SMF= submucous fibroid

0

5000

10000

15000

20000

25000

30000

35000

40000

20 25 30 35 40

Will

ingn

ess

to

pay

(£

/ a

dd

itio

nal

cas

e s

atis

fie

d)

Prevalence of polyp or SMF (%)

OPH and EBx preferred

OPH alone preferred

TVS alone preferred

SIS alone preferred

LNG-IUS alone preferred

206

Deterministic results: Women being managed during multiple clinic

appointments

To reflect ‘traditional’ investigation and treatment of patients, over the course of multiple

clinic appointments, the base case tree was adapted and the results of the deterministic

analysis are displayed in Table 9.4 below.

Table 9.4 Deterministic results of cost-effectiveness analysis for women with heavy

menstrual bleeding managed over multiple clinic appointments


infusion sonography; TVS – transvaginal ultrasound scan

Outcomes

As with the base case analysis, the ‘no investigation’ strategies were the least effective

strategies for managing women. The most effective strategy for investigating women using a

multiple clinic attendance as opposed to a ‘one-stop’ approach was the combination of OPH

+EBx.

Strategy

Cost(£)

Effectiveness

(Satisfaction)


TVS alone 1204 0.9551

EBx alone 1214 0.9460

SIS alone 1217 0.9629

TVS + Ebx 1266 0.9540

SIS + EBx 1274 0.9643

OPH alone 1288 0.9641

TVS + OPH 1315 0.9644

OPH +EBx 1317 0.9674

SIS + OPH 1343 0.9644

TVS+ OPH+EBx 1391 0.9649

SIS +OPH +EBx 1418 0.9650

Hysterectomy alone 3182 0.9335

207

Costs

Costs for the ’no investigation’ strategies (LNG-IUS alone and hysterectomy alone) remained

unchanged. The costs of the investigative strategies however increased due to the additional

appointments required with the costs for the investigation strategies ranging from £1204 to

£1418 in this alternative analysis compared to £1078 to £1256 in the base case analysis.


The strategies "EBx alone", "TVS + EBx", "OPH alone", "SIS + OPH", "TVS + OPH +

EBx", "SIS + OPH + EBx", and "Hysterectomy alone" are excluded by simple dominance and

the strategies "TVS alone", "SIS + EBx" and "TVS + OPH" are excluded by extended

dominance. The remaining three strategies are not dominated. In contrast to the base case

analysis the strategy SIS alone is no longer dominated whereas "OPH alone" is. Table 9.5

displays the non-dominated strategies from deterministic analysis.

Table 9.5 Non-dominated strategies from the analysis of women presenting with heavy


Strategy

Cost

(£)

Incremental

Cost (£) Effectiveness

Incremental

Effectiveness ICER

LNG-IUS alone 1067

0.9333

SIS alone 1217 150 0.9629 0.0296 5070

OPH + EBx 1317 100 0.9674 0.0045 22100 EBx = endometrial biopsy; OPH = outpatient hysteroscopy; SIS = saline infusion sonography; LNG-IUS = levonorgestrel

intrauterine system

Using SIS to investigate women in this strategy costs an additional £5070 to make an extra

women satisfied compared to not investigating and giving all women a LNG-IUS. OPH and

EBx costs an additional £22,100 to the cost of SIS to gain a further satisfied patient. OPH

alone does not appear as a non-dominated option in this analysis. The line on the cost-

effectiveness plane in Figure 9.13 links the non-dominated strategies.

208

Figure 9.13 Cost-effectiveness plane showing the results of deterministic analysis for

strategies to investigate women presenting with heavy menstrual bleeding managed over

multiple clinic appointments. (Hysterectomy removed).

EBx = endometrial biopsy; OPH = outpatient hysteroscopy; SIS = saline infusion sonography; TVS = transvaginal ultrasound

scan; LNG-IUS = levonorgestrel intrauterine system

1000

1050

1100

1150

1200

1250

1300

1350

1400

1450

0.93 0.94 0.95 0.96 0.97

Co

st

Effectiveness

LNG-IUS alone

SIS alone

OPH and EBx

TVS alone

SIS + EBx

TVS + OPH

EBx alone

TVS + EBx

OPH alone

SIS+OPH

TVS+ OPH+EBx

SIS +OPH +EBx

Not Dominated

209

Probabilistic sensitivity analysis results: Women being managed during

multiple clinic appointments

In the cost-effectiveness scatterplot below (Figure 9.14) it is clear from the degree of overlap

of the diagnostic strategies that there is uncertainty regarding which one might be considered

most cost-effective when a range of values is sampled from the distributions of the data

values.


model for each of the individual strategies for investigating women with heavy



infusion sonography; TVS = transvaginal ultrasound scan

0

500

1000

1500

2000

2500

3000

3500

4000

0.8 0.85 0.9 0.95 1

Co

st (

£)

Effectiveness

C(LNG-IUS alone)


C(OPH alone)

C(TVS alone)

C(EBx alone)

C(SIS alone)

C(OPH and EBx)

C(SIS + EBx)

C(SIS + OPH)

C(TVS + Ebx)

C(TVS + OPH)

C(SIS +OPH +EBx)

C(TVS+ OPH+EBx)

210

The CEAF (Figure 9.15) illustrates the overall uncertainty related to the optimal decision

across a range of plausible willingness to pay (WTP) values and shows that up to a WTP

value of approximately £5000 LNG-IUS alone is cost-effective. However between £5000 and

£20,000 SIS alone may be preferable although there is reasonable uncertainty as to whether

this really is the optimal strategy with the probability lying between 40 and 60%. Above

£20,000 OPH and EBx becomes the preferred strategy.


strategies for women with heavy menstrual bleeding managed over multiple clinic

appointments

LNG-IUS = levonorgestrel intrauterine system; EBx = endometrial biopsy; OPH = outpatient hysteroscopy

Probabilistic pair wise comparisons were made between the non-dominated strategies to

explore the uncertainty between them. TVS was also compared to SIS alone because of its

proximity to the line of non-dominance.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P C

ost

-Eff

ect

ive

(O

pti

mal

Str

ate

gy

On

ly)


LNG-IUS alone

SIS alone

OPH and EBx

211

Transvaginal scan versus levonorgestrel intra-uterine system: Women with HMB

managed over multiple clinic appointments

The scatterplot below (Figure 9.16a) shows that the strategy TVS alone is more effective than

LNG-IUS alone and is very likely to be more costly per patient satisfied. The CEAC (Figure

9.16b) shows that above a WTP of £8000 TVS is probably a more cost effective option than

LNG-IUS alone per woman satisfied but it is only at WTP above £20,000 that this is almost

certain (p>0.9).


transvaginal ultrasound scan alone strategy relative to the LNG-IUS alone strategy for

women with heavy menstrual bleeding managed over multiple clinic appointments

-100

-50

0

50

100

150

200

250

300

350

0 0.02 0.04 0.06 0.08

Incr

em

en

tal C

ost

(£

)

Incremental Effectiveness (Satisfaction)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P T

VS

Co

st-E

ffe

ctiv

e


(a) (b)

212

Saline infusion sonography versus levonorgestrel intra-uterine system: Women with

HMB managed over multiple clinic appointments

The scatterplot depicted in Figure 9.17a shows that saline infusion scan is more effective than

LNG-IUS alone but it is also probably more expensive. The CEAC (Figure 9.17b) shows that

at a WTP of around £10,000 it is probable that SIS is the most cost-effective option and this

becomes almost certain at a WTP of £20,000.


saline infusion sonography alone strategy relative to the LNG-IUS alone strategy for

women with heavy menstrual bleeding managed over multiple clinic appointments

-100

-50

0

50

100

150

200

250

300

0 0.05 0.1

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P S

IS C

ost

-Eff

ect

ive


(a) (b)

213

Outpatient hysteroscopy versus saline infusion sonography: Women with HMB

managed over multiple clinic appointments

Figure 9.18a suggests that there is uncertainty between SIS and OPH but that OPH may be

more effective than SIS. It also suggests that OPH is more expensive than SIS. The CEAC

(Figure 9.18b) shows that even at a WTP of £40,000, OPH is unlikely to be the most cost-

effective strategy (p<0.3).


outpatient hysteroscopy alone strategy relative to the saline infusion sonography alone

strategy for women with heavy menstrual bleeding managed over multiple clinic

appointments

-40

-20

0

20

40

60

80

100

120

140

160

-0.005 0 0.005 0.01 0.015 0.02

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P O

PH

Co

st-E

ffe

ctiv

e


(a) (b)

214

Outpatient hysteroscopy and endometrial biopsy versus outpatient hysteroscopy:

Women with HMB managed over multiple clinic appointments

Figure 9.19 shows that OPH and EBx is probably more effective and more expensive than

OPH alone but that it is only becomes likely to be the most cost-effective of the two strategies

at WTP values of above £20,000 and that even at WTP of £40,000 the probability of it being

the most cost-effective strategy is only just above 0.7.


outpatient hysteroscopy with endometrial biopsy strategy relative to the outpatient

hysteroscopy alone strategy for women with heavy menstrual bleeding managed over


-30

-20

-10

0

10

20

30

40

50

60

70

80

-0.005 0 0.005 0.01 0.015

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P O

PH

an

d E

Bx

Co

st-E

ffe

ctiv

e


(a) (b)

215

Saline infusion sonography versus transvaginal scan: Women with HMB managed over

multiple clinic appointments

The scatterplot below (Figure 9.20a) shows that SIS is more effective than TVS although it is

probably a more expensive strategy. The CEAC (Figure 9.20b) shows that SIS is likely to be

the most cost-effective strategy at a WTP of approximately £3000 but this becomes almost

certain (p>0.9) at just over £10,000.


saline infusion sonography alone strategy relative to the transvaginal scan alone strategy

for women with heavy menstrual bleeding managed over multiple clinic appointments.

-80

-60

-40

-20

0

20

40

60

80

100

120

140

0 0.01 0.02 0.03 0.04

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P S

IS C

ost

-Eff

ect

ive


(a) (b)

216

Saline infusion sonography and endometrial biopsy versus saline infusion sonography:


Figure 9.21a shows that SIS + EBx is more effective and more expensive than SIS alone.

Figure 9.21b shows that SIS + EBx is unlikely to be cost-effective at WTP values acceptable

to health service providers.


saline infusion sonography with endometrial biopsy strategy relative to the saline

infusion sonography alone strategy for women with heavy menstrual bleeding managed

over multiple clinic appointments.

-60

-40

-20

0

20

40

60

80

100

120

140

0 0.002 0.004 0.006 0.008

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P S

IS +

EB

x C

ost

-Eff

ect

ive


(a) (b)

217

Outpatient hysteroscopy and endometrial biopsy versus saline infusion sonography and

endometrial biopsy: Women with HMB managed over multiple clinic appointments

The cost-effectiveness plane in Figure 9.22a shows that OPH+EBx is more effective and more

expensive than SIS+EBx. The CEAC (Figure 9.22b) suggests that above £15,000 OPH and

EBx is likely to be the most cost-effective test and that as the WTP increases so does the

certainty.


outpatient hysteroscopy with endometrial biopsy strategy relative to the saline infusion

sonography with endometrial biopsy strategy for women with heavy menstrual bleeding

managed over multiple clinic appointments.

-20

0

20

40

60

80

100

-0.01 0 0.01 0.02 0.03

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P O

PH

+ E

Bx

Co

st-E

ffe

ctiv

e


(a) (b)

218

Outpatient hysteroscopy and endometrial biopsy versus saline infusion sonography:


The scatterplot below (Figure 9.23a) shows that OPH and EBx is probably more effective and

more expensive than SIS. The CEAC (Figure 9.23b) shows that above a WTP threshold of

£25,000 OPH and EBx is likely to be the most cost-effective strategy, although the probability

of this is till only 0.7 at a WTP of £40.000.


outpatient hysteroscopy with endometrial biopsy strategy relative to the saline infusion

sonography alone strategy for women with heavy menstrual bleeding managed over


-50

0

50

100

150

200

-0.01 0 0.01 0.02 0.03

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 20000 40000

P O

PH

+ E

Bx

Co

st-E

ffe

ctiv

e


(a) (b)

219

Deterministic results: Prior treatment with the LNG-IUS

The base case economic analysis assumed that the women presenting to the gynaecologist

with HMB had not had any treatment in primary care. NICE recommend that women with

HMB should be treated with a LNG-IUS in primary care and only be referred to a

gynaecologist if symptoms persist, structural abnormality is expected or contraindications

exist (142). In practice only around 25% of women referred from primary care have received

prior treatment with the LNG-IUS (233). An alternative analysis was performed to examine

the ‘ideal’ scenario, when women have already tried a LNG-IUS, and assess whether the

preferred cost-effective investigative strategies are altered. Disease prevalence was adjusted to

reflect the fact that women treated appropriately with a LNG-IUS in primary care (i.e. without

intracavity pathology, endometrial cancer or large fibroids) were less likely to have persistent

symptoms and need referral to a gynaecologist. It was assumed that fertility was not desired in

this analysis, as in the base case analysis. The strategy LNG-IUS alone could no longer be

used as the reference strategy (now being redundant) and was replaced by a strategy of ‘no

further intervention’ i.e. attending clinic with a LNG-IUS in situ, seeing a gynaecologist but

deciding not to have any further intervention.

220


menstrual bleeding with a LNG-IUS in situ

Strategy Cost (£) Effectiveness

(satisfaction)

No further intervention 1355 0.9039

OPH alone 1681 0.9633

SIS alone 1711 0.9633

TVS + OPH 1746 0.9633

SIS + OPH 1775 0.9633

TVS alone 1785 0.9633

OPH and EBx 1796 0.9628

TVS+ OPH+EBx 1840 0.9628

SIS + EBx 1846 0.9628

SIS +OPH +EBx 1864 0.9629

EBx alone 1942 0.9628

TVS + Ebx 1980 0.9639

Hysterectomy alone 3218 0.9378 LNG-IUS = levonorgestrel intrauterine system; EBx = endometrial biopsy; OPH = outpatient hysteroscopy; SIS = saline


Outcomes

As before, the diagnostic strategies were more effective than the treatment only strategies.

The most effective strategy was combination testing with transvaginal scan and endometrial

biopsy however the difference between all of the diagnostic strategies was minimal, ranging

from 96.28% to 96.39%.

Costs

The results presented in Table 9.6 show that all costs have increased compared with the base

case analysis, reflecting the increased prevalence of organic uterine pathology requiring more

expensive treatments. Adopting no further treatment and persevering with the LNG-IUS

treatment alone (reference strategy) was the cheapest option, costing £1355 per woman

treated for HMB in a secondary care setting. The strategy of hysterectomy for all women was

the most expensive at £3218 per woman treated i.e. £1863 more than the approach of LNG-

221

IUS treatment alone. The cheapest diagnostic testing strategy was the use of outpatient

hysteroscopy alone, costing £1681 for every woman treated, i.e. £326 more than continuation

with LNG-IUS treatment.

Cost-effectiveness & dominance

The testing strategies outpatient hysteroscopy (OPH) alone and transvaginal scan combined

with endometrial biopsy (TVS + EBx) remain non-dominated by alternative empirical

treatment or diagnostic testing strategies. The remaining strategies are all dominated by OPH

alone except for hysterectomy which is dominated by TVS+EBx. Table 9.7 presents the

deterministic analysis restricted to the non-dominated competing strategies.



Strategy

Cost

(£)

Incremental

Cost (£) Effectiveness

Incremental

Effectiveness ICER

No further

intervention 1355

0.9039

OPH alone 1681 326 0.9633 0.0594 5480

TVS + EBx 1980 299 0.9639 0.0006 516000 Apparent anomalies with subtraction are due to rounding effects.

EBx = endometrial biopsy; OPH = outpatient hysteroscopy; TVS=– transvaginal ultrasound scan ; ICER= incremental cost-

effectiveness ratio

222


strategies to investigate women with heavy menstrual bleeding with a LNG-IUS in situ



The line on the graph in Figure 9.24 joins the non-dominated strategies, LNG-IUS only, OPH

alone and TVS and EBx. When LNG-IUS and hysterectomy are removed the relationship of

the other strategies to the line of non-dominance is clearer (Figure 9.25).

0

500

1000

1500

2000

2500

3000

3500

0.9 0.92 0.94 0.96 0.98

Co

st (

£)


Mirena alone

OPH alone

TVS + Ebx

SIS alone

TVS + OPH

SIS + OPH

TVS alone

OPH and EBx

TVS+ OPH+EBx

SIS + EBx

SIS +OPH +EBx

EBx alone

Hysterectomy alone

Not Dominated

LNG-IUS

223


strategies to investigate women with heavy menstrual bleeding with a LNG-IUS in situ.

(Hysterectomy alone and LNG-IUS alone not shown)



1650

1700

1750

1800

1850

1900

1950

2000

0.9625 0.963 0.9635 0.964

Co

st (

£)


OPH alone

TVS + Ebx

SIS alone

TVS + OPH

SIS + OPH

TVS alone

OPH and EBx

TVS+ OPH+EBx

SIS + EBx

SIS +OPH +EBx

EBx alone

Not Dominated

224

Probabilistic sensitivity analysis results: Prior treatment with the LNG-IUS

In the cost-effectiveness scatterplot below (Figure 9.26) it is clear from the degree of overlap

of the diagnostic strategies that there is uncertainty regarding which one might be considered

most cost-effective when a range of values is sampled from the distributions of the data

values.






The CEAF (Figure 9.27) illustrates the overall uncertainty related to the optimal decision

across a range of plausible willingness to pay (WTP) values. The likelihood is that OPH is

cost-effective compared with continuing with the LNG-IUS treatment at WTP thresholds or

0

500

1000

1500

2000

2500

3000

3500

4000

0.7 0.8 0.9 1

Co

st (

£)


CE Scatterplot

C(Mirena alone)


C(OPH alone)

C(TVS alone)

C(EBx alone)

C(SIS alone)

C(OPH and EBx)

C(SIS + EBx)

C(SIS + OPH)

C(TVS + Ebx)

C(TVS + OPH)

C(SIS +OPH +EBx)

C(TVS+ OPH+EBx)

C (LNG-IUS alone)

225

around £20,000. It can be seen that OPH is the preferred option at lower WTP values but that

there is considerable uncertainty as WTP falls below £10,000.


strategies for women with heavy menstrual bleeding with a LNG-IUS in situ

LNG-IUS = levonorgestrel intrauterine system; OPH = outpatient hysteroscopy

Probabilistic pair wise comparisons were made between the non-dominated strategies to

explore the uncertainty between them. SIS was also compared to OPH alone because of its

proximity to the line of non-dominance.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000P C

ost

-Eff

ect

ive

(O

pti

mal

Str

ate

gy O

nly

)


LNG-IUS alone

OPH alone

226

Outpatient hysteroscopy versus levonorgestrel intra-uterine system: Prior treatment

with the LNG-IUS

The scatterplot below (Figure 9.28a) shows that OPH is almost certain to be more effective

than LNG-IUS alone but it will be more costly per patient satisfied. The CEAC (Figure 9.28b)

shows that above a WTP of £6000 OPH is probably a more cost effective option than LNG-

IUS alone per woman satisfied. At a WTP above £15,000 this is almost certain (p>0.9).


Outpatient hysteroscopy alone strategy relative to the LNG-IUS alone strategy for

women with heavy menstrual bleeding with a LNG-IUS in situ

0

100

200

300

400

500

600

700

0 0.05 0.1 0.15 0.2

Incr

em

en

tal C

ost

(£

)

Incremental Effectiveness (satisfaction)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P O

PH

Co

st-E

ffe

ctiv

e


(a) (b)

227

Transvaginal ultrasound scan and endometrial biopsy versus outpatient hysteroscopy:

Prior treatment with the LNG-IUS

The cost-effectiveness plane, Figure 9.29a shows that there is considerable overlap between

‘TVS and EBx’ and ‘OPH alone’. It certainly is not clear which is the most effective strategy

but TVS and EBx is the more costly of the two. For all WTP values between £0 and £40,000

per extra woman satisfied, OPH alone is almost definitely the most cost-effective option when

compared to TVS and EBx (p>0.98) as displayed in the CEAC below (Figure 9.29b)


Transvaginal scan and endometrial biopsy strategy relative to the outpatient

hysteroscopy alone strategy for women with heavy menstrual bleeding with a LNG-IUS

in situ

0

100

200

300

400

500

600

700

-0.02 -0.01 0 0.01 0.02 0.03

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 20000 40000

P T

VS

+ EB

x C

ost

-Eff

ect

ive


(a) (b)

228

Saline infusion sonography versus outpatient hysteroscopy: Prior treatment with the

LNG-IUS

SIS falls close to the line of non-dominance in the deterministic analysis therefore a

comparison was made between SIS and the most cost-effective, non-dominated strategy, OPH

alone. The graph, Figure 9.30a, shows that although there is likely to be little difference in

effectiveness between the two strategies, SIS is almost certain to be the more costly strategy.

The CEAC (Figure 9.30b) comparing SIS and OPH shows that SIS is extremely unlikely to be

the most cost-effective option even at a WTP of £40,000 per additional woman satisfied.


Saline infusion scan alone strategy relative to the outpatient hysteroscopy alone strategy

for women with heavy menstrual bleeding with a LNG-IUS in situ

-40

-20

0

20

40

60

80

100

120

-0.004 -0.002 0 0.002 0.004

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P S

IS C

ost

-Eff

ect

ive


(a) (b)

229

Deterministic analysis: Women wishing to preserve their fertility

Table 9.8 shows the deterministic analysis of the twelve strategies following re-configuration

of the decision trees to reflect women who would not have a hysterectomy or an endometrial

ablation to treat their HMB because they wished to maintain their long term fertility.


menstrual bleeding who wish to preserve their fertility

Strategy

Cost (£)

Effectiveness

(satisfaction)


SIS alone 800 0.8467

OPH alone 844 0.8629

SIS + OPH 944 0.8649

TVS alone 740 0.8033

EBx alone 754 0.7419

TVS + Ebx 870 0.8020

TVS + OPH 913 0.8389

OPH and EBx 914 0.8618

SIS + EBx 971 0.8473

TVS+ OPH+EBx 971 0.8572

SIS +OPH +EBx 1003 0.8584 LNG-IUS = levonorgestrel intrauterine system; EBx = endometrial biopsy; OPH = outpatient hysteroscopy; SIS = saline


Outcomes

Satisfaction rates are reduced when compared to the original analysis as optimal surgical

interventions precluding future fertility are not available for women with large fibroids or

those resistant to the LNG-IUS. This is reflected by the lower satisfaction rate of 65.6% for

this analysis compared with 93.33% in the original when women receive LNG-IUS without

any investigation. Satisfaction rates for the investigative strategies range from 74.19% for

TVS to 86.49% for SIS and OPH. There is greater variation between the satisfaction rates in

this analysis than in the base case when values varied marginally between 94.6% and 96.7%.

230

Costs

It should be noted that the costs decreased when compared to the original analysis. Cost is

decreased because the more expensive treatments tend to be the surgical options (endometrial

ablation and hysterectomy) which are contraindicated in women desiring preservation of their

fertility. The cost of LNG-IUS alone has decreased to £421 from £1066 in the base case

analysis because the women identified to have large fibroids do not undergo a hysterectomy

and women who are dissatisfied with the LNG-IUS cannot be offered any further treatment.

The cheapest investigative strategy in this analysis is TVS alone, costing £740 per patient.

The most expensive strategy is the combination of SIS, OPH and EBx costing £1003 per

patient.


The strategy of EBx alone is dominated by TVS alone ,which in turn is dominated by

extended dominance due to a blend of LNG-IUS alone and SIS alone. The remaining

strategies are dominated by either OPH alone or SIS and OPH together. Once the dominated

strategies were removed the testing strategies which remained were SIS alone, OPH alone and

SIS and OPH together. This can be more clearly appreciated in Table 9.9. In contrast to the

base case analysis, a combination strategy of OPH + EBx is not potentially cost-effective.

Moreover, SIS alone or in combination with OPH is non-dominated whereas in women

without the need to preserve their fertility (base case), SIS and related strategies were not

cost-effective.

231



Strategy

Cost

(£)

Incremental

cost (£)

Effectiveness

(satisfaction)

Incremental

effectiveness ICER

LNG-IUS

alone 421

0.6557

SIS alone 800 378 0.8467 0.1910 1980

OPH alone 844 44 0.8629 0.0162 2720

SIS + OPH 955 100 0.8649 0.0020 50300 Apparent anomalies with subtraction are due to rounding effects.

LNG-IUS = levonorgestrel intrauterine system; OPH = outpatient hysteroscopy; SIS = saline infusion sonography; ICER=

incremental cost-effectiveness ratio.

Figure 9.31 Total costs and effectiveness of the alternative strategies for the diagnostic

work up of HMB for women wishing to preserve their fertility. (Hysterectomy has been

excluded).



The line on the cost-effectiveness plane (Figure 9.31) joins the non-dominated strategies,

starting with the base case of LNG-IUS which joins to SIS, followed by OPH and then SIS

0

200

400

600

800

1000

1200

0.6 0.7 0.8 0.9

Co

st (

£)


Deterministic Analysis preserving fertility

Mirena alone

SIS alone

OPH alone

SIS + OPH

TVS alone

EBx alone

TVS + Ebx

TVS + OPH

OPH and EBx

SIS + EBx

TVS+ OPH+EBx

SIS +OPH +EBx

Not Dominated

LNG-IUS alone

232

and OPH. TVS lies close to this line and therefore when exploring the results TVS was

included to see whether analysing the spread of results might suggest that TVS could become

cost-effective when values are varied around their point estimates.

233

Probabilistic sensitivity analysis results: Women wishing to preserve their

fertility

Figure 9.32 shows the uncertainty around the absolute cost and effectiveness values for each

of the strategies. Hysterectomy alone has been removed as it is too expensive to be a

competing strategy and removing it allows clearer presentation of the other strategies. There

is overlap between strategies.






0

200

400

600

800

1000

1200

1400

0.5 0.6 0.7 0.8 0.9 1

Co

st (

£)


CE Scatterplot

C(Mirena alone)

C(OPH alone)

C(TVS alone)

C(EBx alone)

C(SIS alone)

C(OPH and EBx)

C(SIS + EBx)

C(SIS + OPH)

C(TVS + Ebx)

C(TVS + OPH)

C(SIS +OPH +EBx)

C(TVS+ OPH+EBx)

C(LNG-IUS alone)

234

The acceptability frontier in Figure 9.33 shows how likely the non-dominated strategies are to

be the most cost-effective option at a range of WTP thresholds. The strategy SIS and OPH is

not plotted on the CEAF because it only becomes cost-effective at a WTP too high to be

acceptable to the NHS.

Figure 9.33 Cost-effectiveness acceptability frontier showing the preferred diagnostic

strategy over a range of WTP thresholds for women who wish to preserve their fertility

LNG-IUS = levonorgestrel intrauterine system; SIS = saline infusion sonography OPH = outpatient hysteroscopy.

The uncertainty represented in the CEAF is explored by considering pair wise comparisons

between the adjacent non-dominated options. TVS is also explored because its mean is close

to the non-dominance line.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000P C

ost

-Eff

ect

ive

(O

pti

mal

Str

ate

gy O

nly

)


Acceptability Frontier

LNG-IUS alone

SIS alone

OPH alone

235

SIS alone vs. LNG-IUS alone: Women wishing to preserve their fertility

Figure 9.34 shows that SIS is definitely more effective and more costly than LNG-IUS alone.

The CEAC (Figure 9.34b) shows that at above a WTP of £2700 SIS is definitely (p=1) a more

cost-effective option than using LNG-IUS alone.


Saline infusion sonography alone strategy relative to the LNG-IUS alone strategy for

women wishing to preserve their fertility

0

100

200

300

400

500

600

0 0.1 0.2 0.3

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P S

IS a

lon

e C

ost

-Eff

ect

ive


(a) (b)

236

Outpatient hysteroscopy versus saline infusion: Women wishing to preserve their

fertility

Figure 9.35a shows that outpatient hysteroscopy is likely to be more effective and more

expensive than saline infusion scan for investigating women with HMB who want to preserve

their fertility. The CEAC (Figure 9.35b) shows that above a WTP of approximately £2500 to

make one extra woman satisfied, OPH becomes likely to be the most cost effective strategy

(p>0.5) and that at a WTP just above £5000 it has a greater than 90% chance of being the

most cost-effective strategy when compared to SIS.


Outpatient hysteroscopy alone strategy relative to the saline infusion sonography alone

strategy for women wishing to preserve their fertility

-50

0

50

100

150

200

-0.02 0 0.02 0.04 0.06 0.08

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P O

PH

alo

ne

Co

st-E

ffe

ctiv

e


(a) (b)

237

Saline infusion sonography and outpatient hysteroscopy versus outpatient hysteroscopy:

Women wishing to preserve their fertility

Figure 9.36a shows the cost-effectiveness plane comparing OPH with SIS to OPH alone. It

shows that SIS with OPH is more costly than OPH alone and that this extra cost is almost

certainly between £100 and £120. The two test combination is also probably but not

definitely, the more effective strategy. The CEAC (Figure 9.36b) shows the proportion of

model replications for which saline infusion scan and outpatient hysteroscopy is preferred to

outpatient hysteroscopy alone at any given willingness to pay per additional case satisfied. It

shows that even at a WTP of £40,000 there is considerable uncertainty that SIS and OPH is a

more-cost-effective strategy than OPH alone (p=0.2).


Saline infusion sonography and outpatient hysteroscopy strategy relative to the

outpatient hysteroscopy alone strategy for women wishing to preserve their fertility

0

20

40

60

80

100

120

140

160

-0.005 0 0.005 0.01

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P S

IS +

OP

H C

ost

-Eff

ect

ive


(a) (b)

238

Transvaginal ultrasound scan versus levonorgestrel intra-uterine system: Women

wishing to preserve their fertility their fertility


TVS alone and LNG-IUS. It shows that TVS is more effective and more costly than giving all

women an LNG-IUS without investigation. The CEAC (Figure 9.37b) shows the proportion

of model replications for which transvaginal scan is preferred to LNG-IUS alone at any given

willingness to pay per additional case satisfied. It shows that above a WTP of £3,000 TVS is

definitely the most cost-effective option (p=1).


Transvaginal ultrasound scan alone strategy relative to the LNG-IUS alone strategy for

women wishing to preserve their fertility

0

100

200

300

400

500

600

0 0.1 0.2 0.3

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P T

VS

alo

ne

Co

st-E

ffe

ctiv

e

Willingness to Pay (£/ additional case satisfied)

(a) (b)

239

Saline infusion sonography versus transvaginal ultrasound scan: Women wishing to



SIS alone and TVS alone. It shows that SIS is probably more effective and more costly that

TVS alone. The CEAC (Figure 9.38b) shows the proportion of model replications for which

SIS is preferred to TVS alone at any given willingness to pay per additional case satisfied. It

shows that above a WTP of approximately £4000 SIS alone is definitely the most cost-

effective option when compared to TVS.


Saline infusion sonography alone strategy relative to the transvaginal ultrasound scan


-150

-100

-50

0

50

100

150

200

-0.05 0 0.05 0.1

Incr

em

en

tal C

ost

(£

)


0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 10000 20000 30000 40000

P S

IS a

lon

e C

ost

-Eff

ect

ive


(a) (b)

240

CHAPTER 10

DISCUSSION OF THE ECONOMIC ANALYSIS OF



Discussion of the main findings

The objective of this economic analysis was to evaluate the cost-effectiveness of diagnostic

testing strategies for HMB. Universal treatment with the LNG-IUS without any preliminary

investigation was chosen as the reference strategy to compare testing options against. This

was because (i) not investigating or treating is an unacceptable alternative; (ii) the LNG-IUS

is recommended by NICE as first line treatment of HMB. Treating all women with a LNG-

IUS without preliminary testing resulted in high levels of satisfaction (93%) which were

increased by approximately 4% if some form of currently available diagnostic testing was

undertaken to guide treatment.

Base Case

The economic model which adopted a ‘one-stop’ setting for the base-case analysis reflects

contemporary practice and will remain relevant as services evolve.

Two potentially cost-effective investigation strategies were identified; initial testing with

outpatient hysteroscopy alone (OPH) or a combination strategy incorporating outpatient

hysteroscopy with endometrial biopsy (OPH + EBx). Although the strategy OPH +EBx was

241

marginally more effective, the incremental cost-effectiveness ratio (ICER) was approximately

£21,000 to gain one more satisfied patient compared to OPH, while OPH has an ICER of

under £400 per additional case satisfied compared to LNG-IUS alone. Thus for relatively little

additional investment by the NHS, the adoption of OPH in place of LNS-IUS alone will

improve outcomes for women presenting with HMB. This improvement can be increased

further if combination testing with OPH + EBx is introduced. This additional cost is

contentious, but it can be tested by comparison to the £20,000-£30,000 per QALY NICE

threshold at which interventions are considered cost-effective to implement within the NHS.

Due to paucity of HRQL data, QALYs could not be used for the cost effectiveness analysis

however an estimation can be made to give an approximate cost per QALY. It has been

estimated that a woman’s quality of life is reduced by 0.5 for the one week per month of

heavy menses (234). This means that HMB is associated with a reduction of 0.125 QALY in

any year, as the reduction of 0.5 applies for one quarter of the time overall. Considering a 45

year old woman with seven symptomatic years until menopause, future years need to be

discounted at 3.5%, so that the annual QALY loss must be multiplied by [1 + (1/1.035) +

(1/1.035)2 + (1/1.035)

3 + ... + (1/1.035)

6 ] or approximately 6.3 to give a total QALY loss of

approximately 0.8 QALY. This means that an ICER of £21,000 per case satisfied is

approximately equivalent to an ICER of £26,500/QALY, which falls within the £20,000-

£30,000 per QALY threshold range used by NICE. Therefore, the strategy OPH+EBx is of

borderline cost-effectiveness compared to initial investigation with OPH in isolation.

The certainty of the base-case results was assessed by probabilistic sensitivity analysis.

Outpatient hysteroscopy remained the more cost-effective strategy even at relatively low

willingness to pay thresholds. Outpatient hysteroscopy with EBx also remained stable

however, to be at least 70% certain that it was a more cost-effective alternative than OPH

242

alone, the WTP threshold would need to be increased to around £40,000 per patient satisfied

(~ £50,000 / QALY). Endometrial biopsy is most useful for diagnosing endometrial disease

but the prevalence of endometrial hyperplasia and carcinoma is low in a premenopausal

population estimated at around 5%. In populations of women where higher rates of

endometrial disease are observed (e.g. epidemics of obesity), the benefit of EBx will have

more influence on overall cost-effectiveness. However, substantially higher estimates of

endometrial disease in an HMB population are unlikely.

Ultrasound is a convenient, minimally invasive, portable test that allows assessment of both

the uterus and ovaries. It is universally available and can be incorporated easily into standard

gynaecological examination. In view of the popularity of pelvic scanning and the fact that two

scanning strategies came close to the boundary of dominance when the cost-effectiveness

analysis was performed, further exploratory analyses were undertaken. Sensitivity analysis

was performed to examine transvaginal ultrasound (TVS) and saline infusion sonography

(SIS) to see whether there was uncertainty regarding them when compared to the most cost-

effective strategy of OPH alone. However when compared to OPH, TVS was found to be

almost certainly less effective and there was also a chance of it being more costly too

therefore ruling it out as a primary diagnostic test. When SIS was compared to OPH there was

significant doubt regarding which was the most effective approach. By increasing the

willingness to pay threshold from £0 to £40,000, the likelihood that OPH was more cost

effective than SIS increased further from 60% to 80%. Thus, at WTP levels acceptable to the

NHS, OPH appears preferable to SIS as a first-line diagnostic test in HMB. However,

clinicians who currently use SIS for first line investigation of HMB and are able to easily

integrate SIS into their practice at low cost, may have less to gain by changing to OPH.

243

Furthermore, patients receive added reassurance from knowing that their ovaries have been

examined as well as the uterine cavity, a benefit of scanning that is neglected by OPH.

In sensitivity analysis, the cost of SIS needed to be reduced from £71 to £52 to make it more

cost-effective than OPH. This reduction in costs is unrealistic as SIS would have to cost less

than the £55 which is the cost of a standard 2D transvaginal scan which is unfeasible given

that it requires additional equipment (instillation catheter) and takes longer to perform; costs

which have been estimated to be an extra 35% on top of transvaginal scan in a Dutch study

(156).

Prevalence of focal uterine pathology

Sensitivity analyses were conducted to investigate whether the prevalence of polyps/SMFs

could have biased the results towards OPH. The prevalence was reduced sequentially to

determine at what prevalence an alternative option may be favoured. This analysis suggested

that even if the estimate of polyps/SMFs was 10% less than had been stated in the base case

(i.e. 28% instead of 38%), OPH would still remain the preferred option. The prevalence of

polyps and submucous fibroids reported in the better quality prevalence studies used in the

model reported values for both pathologies of around 20% (202;205), which approximates to

the 40% prevalence quoted by NICE in their HMB guideline (30% submucous fibroids and

10% endometrial polyps) (142), so if 40% is a true reflection of disease then OPH is for

certain the most cost-effective option. Even in different HMB populations it seems unlikely

that the combined prevalence of endometrial polyps and submucous fibroids would be lower

than 28%, thus there is reasonable certainty that OPH is the most cost-effective diagnostic

strategy in HMB, even in populations with focal uterine pathology prevalence rates at the

lower end of the plausible range.

244

Women managed over multiple clinic appointments

Whilst one-stop services are being increasingly developed, the traditional multi-stop model of

care remains embedded in much of the NHS, therefore an adapted version of the model

investigated the multi-stop approach to investigation and treatment of women with HMB.

This analysis suggested that SIS alone, at an added cost of £5070 per additional patient

satisfied or OPH and EBx at an added cost of £22,100 were the most cost effective strategies.

OPH alone was not considered a cost-effective option in this analysis, however, costs were

applied to OPH that made it less competitive than the other tests by dictating that concomitant

treatment of polyps and fibroids could not be performed at the time of diagnostic

hysteroscopy. This approach is practised at some centres across the UK to try and avoid

delays in clinic running time, so this ‘worst-case scenario’ was used in the multi-stop analysis.

OPH become relatively expensive because it required an additional consultant appointment

whereas TVS and SIS were performed in the scan department by a sonographer and

endometrial biopsy was performed at the initial consultant GOPD appointment. Interestingly,

in the pairwise comparison between OPH and SIS, OPH was probably more effective than

SIS but it was more expensive under the assumptions of the multi-stop model of care to a

degree that OPH was no longer cost-effective compared with SIS. In both the one-stop base

case model of care and the multi-stop reanalysis, the testing strategy of OPH combined with

EBx was a non-dominated strategy. If the same QALY rules are applied to this strategy as to

the base case results (see the discussion under ‘Base case’ above) the cost per QALY for

OPH+EBx was £27,625 which falls within the NICE threshold of £20,000 to £30,000. Thus,

an initial testing approach for women presenting with HMB using OPH with EBx is

potentially cost-effective regardless of the model of care. However, those able to provide a

contemporary ‘one-stop’ service would need to weigh up the additional costs associated with

245

OPH+EBx over simply OPH alone to gain a further health benefit. Similarly, those Units

utilising a traditional multi-stop set up would need to consider the much reduced costs of

initial testing with SIS against the reduction in health benefit when compared to a

combination of OPH + EBx.

It seems likely that the drive to provide outpatient testing and where possible simple

concomitant treatments will continue in light of the convenience for patients and their doctors

as well as the on-going improvement in health technologies. Therefore, because the base case

analysis based upon the premise of a one-stop service, indicated that OPH was the optimal

strategy (regardless of wish for fertility or prior treatment with LNG-IUS (see below)), it may

be consistent to employ combination testing from the outset with OPH and EBx in hospitals

that don’t yet have ‘one-stop’ facilities in place. This would allow the services in these centres

to evolve until they were able to offer ‘one-stop’ care based upon initial testing with OPH in

women presenting with HMB. The alternative for these hospitals would be to adopt SIS as the

first line test for investigating HMB, training doctors, nurses and sonographers in the

technique. However, future re-investment in equipment and training to establish one-stop

services based upon initial testing with OPH may become necessary if such service models

become embedded.

Fertility preservation

Heavy menstrual bleeding is most common in parous women in their fifth decade of life

(186). Thus, the majority of women presenting to secondary care have completed their

families. However, some women with HMB do want to retain their fertility potential and to

take account of this and to test the generalisability of the base case findings an additional

246

model was produced for women wishing to preserve their fertility which precluded certain

surgical treatment options. The findings of this additional analysis were consistent with the

base case scenario with OPH remaining the most cost-effective first line diagnostic test of

choice. SIS was the closest, viable competing testing option. Adopting OPH rather than SIS,

costs an additional £2720 to achieve an extra woman satisfied but this is likely to be

considered affordable and worthwhile by health services. Moreover, at willingness to pay

(WTP) thresholds of around £5000, there is a greater than 90% certainty of OPH being the

most cost-effective option. Indeed, SIS only appeared to be cost-effective over a very narrow

range of WTP values and even then there was considerable observed uncertainty.

Universal LNG-IUS treatment prior to referral

A further analysis using a modified decision model, was based upon the situation envisaged

by NICE in their 2007 report into the management of HMB (142) where all women referred

from primary care with HMB have received prior treatment with a LNG-IUS. Even in this

scenario, OPH continued to be the favoured, cost-effective option with an ICER of £5480 for

each additional woman satisfied and increasing certainty of cost-effectiveness with

increasing, but viable, WTP levels. Although the combination of TVS and EBx was a more

effective approach, it’s ICER of over £500,000 prohibits its use as a sensible diagnostic

strategy for adoption by the NHS.

‘See and Treat’

Outpatient hysteroscopy is consistently the preferred first line testing strategy for women with

HMB irrespective of their desire for future fertility and regardless of pre-referral treatment

with the LNG-IUS. One consideration driving the economic benefits of OPH over other tests

247

is that the modality allows treatment for most pathology to be initiated at the same time as the

diagnosis, thereby reducing the number of patient attendances and costs. This contemporary

’see & treat’ ambulatory approach in gynaecological practice is increasingly being adopted.

There are also additional benefits of embracing this philosophy not accounted for in the

analysis e.g. improved safety, lowered infection risks, convenience, rapid discharge and

recovery (6). Only 30% of women in the model required a further appointment in order to

return for treatment at a later stage in the OPH testing strategy. This was because all women

presenting with polyps or pathology suitable for treatment with a LNG-IUS could be treated

at their first appointment if diagnosed correctly. For scanning modalities, concomitant

treatment was still possible for the same proportion of women, however the cost is elevated

slightly because of the additional cost of hysteroscopy and polypectomy for the 19% of

women with endometrial polyps. Furthermore, the accuracy of TVS for the diagnosis of focal

pathology is reduced compared with OPH. This consideration combined with the additional

costs of the OPH needed to perform the polypectomy for both TVS and SIS scanning

approaches, contributes to their reduced cost-effectiveness compared with OPH. This was

even more apparent for strategic testing approaches utilising EBx from the outset where no

simultaneous treatment was possible because of the necessity to await the EBx result before

making a diagnosis and instituting therapy. ‘Real time’ bedside testing can be seen to confer

an advantage in the efficiency of delivering care. Thus, it is the ability to concomitantly treat

as well as diagnose intrauterine pathology with a high degree of accuracy that leads to OPH

being the most cost-effective strategy testing strategy in HMB.

248

Strengths and limitations

The economic evaluation was based upon the construction of comprehensive, contemporary

and clinically informed decision trees. All available testing strategies were modelled and a

modern, ‘one-stop’ clinical environment for testing assumed. Test accuracy and performance

data were mainly obtained from high quality systematic quantitative reviews of the literature.

Other clinical parameter inputs, including treatment effectiveness and disease prevalence were

obtained following systematic searches and selection based upon a rigorous evaluation of data

quality. Data for the prevalence of fibroids were taken from a database of women from

Birmingham and the surrounding area. It should be recognised that this may not be a true

reflection of the UK prevalence of fibroids due to the large number of women from ethnic

minorities. Producing a comprehensive series of reviews was not the aim of the economic

modelling study so although data were identified using a systematic approach, ‘systematic

reviews’ per se were not conducted for each clinical question. When data were not available

in the published literature or there was uncertainty regarding the assumed diagnosis or need

for further testing, a panel of senior gynaecologists was consulted.

The economic evaluation took the UK NHS as its perspective. This meant that only costs

incurred by the NHS were included, and benefits were measured in patient satisfaction. The

prime aim of testing in chronic benign conditions like HMB is to improve patient

symptomatic outcomes. Thus the approach of evaluating effectiveness in terms of patient

satisfaction within the HMB analysis was relevant. Patient satisfaction was chosen in order to

(i) ensure consistency with previous research trials undertaken in the field (136) and (ii)

because patient satisfaction is deemed to be the co-primary measure of importance (together

249

with menstrual bleeding) by the Cochrane Menstrual Disorders & Subfertility Group for

reviews of interventions for heavy menstrual bleeding (193). Inevitably, this perspective will

have excluded other potentially important costs and benefits. For example, the scarcity of

HRQL data in AUB precluded a cost-utility analysis and comparison by health care decision

makers with competing health care interventions. Whilst feasibility of testing was taken into

account, the morbidity (anxiety, discomfort, complications) and psychological implications

for women and their families was not. Similarly patient preference for testing was not

incorporated into the models or the added value of individual tests outside of the focus of

uterine assessment e.g. simultaneous assessment of the ovaries.

The findings of this analysis appear generalisable to modern practice in the UK because a

contemporary, one-stop clinical setting was used and it was assumed that minor therapeutic

interventions were implemented during the initial visit; when appropriate. Costs were derived

from up-to-date HRG data (235); the cost of individual tests were derived and added to the

cost of a standard new consultation. This method allowed different strategies to be compared

fairly by breaking down the different aspects of each appointment. However, whilst this

approach allowed costs in a contemporary ‘one stop’ clinical setting to be applied, the costs

may not accurately reflect the true cost to the NHS. It is unlikely that the HRG codes

precisely take account of costs from the clinician’s perspective i.e. the additional time and

disruption associated with combination testing either from the outset or conditional on

preceding tests. For example, the ‘one stop’ diagnostic testing philosophy combined with the

ability to provide concomitant treatment (‘see & treat’) is certainly convenient and efficient

but the additional time required (which is somewhat unpredictable) necessitates the allocation

of fewer appointment slots within such clinics. In sensitivity analysis the base case model was

250

adapted to reflect multi-stop diagnostic and treatment pathways to help inform health

providers currently using traditional service models.

Whilst it is certainly the case that a one-stop approach is efficient, obviating the need for

further appointments, the UK NHS perspective of this economic analysis may have

underestimated the total cost savings of the one-stop service. If a wider social perspective was

adopted, the economic benefits of minimising interference to women’s lives by avoiding

unnecessary follow up appointments and the consequent travel and waiting times, would

probably have been substantial. If the costs of absence from paid work and household

activities had been quantifiable in monetary units, they may not have been insignificant. Thus

it is probable that a one-stop setting for investigation of HMB from a societal viewpoint

would have been even more cost-effective compared with traditional multi-stop models of

service delivery.

A variety of different benign pathologies underlie HMB and each pathology has its optimal

first-line treatment. However, some pathologies e.g. DUB and endometrial hyperplasia, share

the same ‘optimal’ first-line treatment i.e. the LNG-IUS. Moreover, erroneously diagnosed

pathologies given ‘incorrect’ treatment, will generally, still respond to several treatments

albeit to a suboptimal degree. Thus, the fact that HMB is usually of benign origin and often

responds to treatment with the LNG-IUS to some degree, regardless of pathology, has raised

uncertainty over the cost-effectiveness of any testing in HMB. It was for this reason that ‘no

investigation’ but LNG-IUS treatment for all was used as the reference strategy against which

all other testing options could be compared. However, the universal prescription of the LNG-

IUS in secondary care to treat HMB is unlikely to be acceptable to either women who require

251

an explanation for their symptoms (and may not want a LNG-IUS although the model did not

account for patient preferences) or clinicians who are aware that more effective, tailored

treatments are available according to the underlying diagnosis. In primary care the situation is

different with NICE recommending LNG-IUS as a first-line treatment without stipulating the

need for diagnostic testing (142). In view of this, a scenario was modelled where all women

with HMB referred to secondary care were refractory to treatment with the LNG-IUS. The

inference from this alternate analysis, that investigation should be based upon first-line testing

with OPH, was unchanged. Thus, given the data available, the model is comprehensive,

pragmatic and relevant to current clinical practice.

Extensive probabilistic sensitivity analyses were conducted to allow for uncertainty,

manipulating parameter inputs for clinical assumptions (disease prevalence, test performance,

accuracy, and consequences of false diagnoses and effectiveness of treatments) and costs.

Acceptability frontiers were drawn to aid assessment and interpretation. The most cost-

effective testing strategies remained stable following sensitivity analyses increasing

confidence in the conclusions To further evaluate the stability and also transferability of the

findings in HMB, alternative models were produced to embrace scenarios where referral to

secondary care is restricted to women who remain refractory to effective medical treatment

(142) and where preservation of fertility is required.

Simplifications

The aim was to develop economic models that accurately and explicitly mirrored clinical

practice. Some simplifications were necessary, driven by a desire to keep the extensive and

comprehensive decision trees workable. One problem was how to account for failed testing

252

and discordance of tests within combination strategies. The accuracy of individual outpatient

tests varies according to which pathology is under scrutiny. However, this does not mean that

combination testing will be more cost-effective. The model comprehensively evaluated tests

used in isolation (with additional testing conditional on the test outcome) and in combination

from the outset (again with any further testing being conditional on the combination testing

outcome). However, when combination testing was incomplete due to tests within the

combination failing, clinicians would then face a dilemma as to what to do next .It was

decided to simplify the model with the aim of maintaining consistency and minimising

potential bias. Thus the approach used was that any test or tests that failed within a

combination testing strategy were assumed to have not allowed a diagnosis to be arrived at

(i.e. the testing strategy has failed) and recourse to hysteroscopy, D&C was the result. The

limitation of this approach, is its clinical validity and relevance. Undoubtedly many clinicians

would consider examination under anaesthesia with hysteroscopy D&C as the ‘gold standard’

test whereas others may pragmatically arrive at a diagnosis based upon the information

provided by the successful tests within the combination strategy. However, there are a huge

amount of potential test combinations and failure possibilities to consider. Moreover, it was

hard to arrive at any consensus from the expert panel as to what subsequent testing or

management decisions would then arise. Therefore the simplification of the model, such that a

combination testing strategy was considered incomplete necessitating recourse to D&C, was

reasonable in order to maintain consistency and minimise bias. Moreover, the failure rates of

tests are generally low and so the impact of this model simplification should have been

minimal.

253

Subtle differences between testing strategies may have been overlooked, for example it was

assumed that women could only have one underlying pathology whereas in practice they may

co-exist. The most common pathologies underlying HMB were accounted for in the model;

DUB, polyps/SMFs, small and large fibroids and significant endometrial disease.

Adenomyosis, a condition of the myometrium (uterine muscle) where ectopic endometrial

tissue is found, can be associated with HMB but was excluded from the model as it tends to

present with pain rather than HMB.

All relevant, widely employed testing modalities were modelled. Endometrial tissue sampling

was assumed to be by outpatient EBx and inpatient D&C in failed cases. Endometrial biopsy

can however be performed under direct vision by passing miniature forceps down the working

channel of an outpatient hysteroscope. This method was not included as it is not widely used

and is less likely to provide an adequate tissue sample for histological assessment compared

with simpler, cheaper and universal outpatient EBx. Furthermore, OPH and SIS are accurate

tests for the diagnosis of focal pathology without recourse to histological confirmation. In

addition, focal endometrial disease (cancer and hyperplasia) necessitating a directed biopsy is

rare with most endometrial conditions being global hormonally induced phenomena allowing

‘blind’ sampling technologies to obtain representative, diagnostic samples with a high level of

accuracy (150;151).

Testing with magnetic resonance imaging was not examined because NICE recommend that it

should not be used for diagnosis in women with HMB on the basis that there are no studies

examining its use (142). One study compared MRI to TVS for identification of adenomyosis

and found no significant difference between them (236). In addition, MRI is not suitable for

254

real time outpatient (‘office’) testing and is expensive and rarely used in the diagnosis of

HMB unless there are concerns over the nature of uterine fibroids (an infrequent situation) or

when suitability for uterine artery embolisation is being assessed.

Assumptions

As with all economic modelling exercises, assumptions had to be made when contentious

clinical decision making or paucity of clinical data were encountered and these assumptions

were ultimately endorsed by the expert clinical panel. One such area pertained to disease

prevalence, estimates of which came from a variety of sources and this meant that to make

them sum to one, the prevalence of one ‘disease’, DUB, needed to be a flexible value. The

assigned value of 0.32 (32%) may not accurately reflect the true prevalence. However,

published data and the expert panel considered this a reasonable estimate. As with the

precision of all assumptions, they were subject to sensitivity analyses. In this particular case,

the inferences of the base case analysis were not affected with varied, plausible estimates of

DUB prevalence. A similar problem was encountered for some accuracy data such that the

reciprocal false positive rate estimates, which had originated from diverse data sources, had to

be manipulated to total one and these were tested in sensitivity analysis.

The selected outcome measure was satisfaction with treatment at one year because this is one

of the most common outcomes measured in RCTs of interventions for HMB (184) and is

clinically relevant. Moreover, the availability of other patient reported outcome measures e.g.

health related quality of life data, (and especially their application to the specific scenarios

that arose from the construction of the comprehensive decision trees) are scarce.

Approximating data from published studies was decided to be an inaccurate method for

255

producing QALYs and so ‘cost per patient satisfied’ was used alone. In addition, the cyclical

and intermittent nature of HMB makes it somewhat problematic to calculate QALY’s

particularly as women are only affected by the condition for approximately 25% of the year

and symptoms will naturally resolve in time once the menopause is reached. Systematic

searching and quality appraisal were used to try and identify the optimal data, but with some

of the interventions used for particular underlying diagnoses, satisfaction data were not

explicitly available. In these cases it was assumed that the outcomes reported (e.g. “reduced

bleeding”, “would recommend the treatment”, “cured of cancer”) were indirect measures of

satisfaction.

Occasionally, satisfaction data were not reported at one year in which case the data collected

closest to one year were used. Using ‘one year’, whilst reasonable in terms of evaluating

medium term response, may favour conservative, ‘uterine-sparing’ procedures (the LNG-IUS,

endometrial ablation and myomectomy) because their effectiveness reduces over time (136).

In contrast, more invasive but definitive hysterectomy is not associated with recurrence of

HMB symptoms. Thus, longer term outcome assessment may have made the option of

‘hysterectomy without diagnostic testing’ a more viable option. However, the objective and

emphasis of the analysis was to estimate the cost-effectiveness of diagnostic testing. Thus, as

long as the treatment options were appropriate and consistently applied according to the

diagnosed pathology, the most cost-effective testing strategies should have been delineated.

The base case analysis assumed included women to be 45 years old as this is the modal age

for presentation to secondary care with HMB. Thus they could be expected to have an average

of a further seven years of menstruation before the menopause.

256

The ability of OPH to facilitate the removal of uterine polyps in addition to simply providing

a diagnosis has been highlighted. However, this presumed cost-efficient benefit may be

blunted if uterine polyps are not causative of HMB, but simply an incidental finding. NICE

were unable to find any data to link uterine polyps and HMB (142) although clinicians do

assume some link (172;237). Two systematic reviews of generally low quality, observational

studies do however, support the notion of polyp treatment being associated with a 75-100%

improvement in HMB complaints (166;167). Even if polyps don’t cause HMB they may

negatively affect treatment (e.g. impairing the effectiveness of the LNG-IUS) and ultimately

cause dissatisfaction (238;239). This may also be a psychological effect given that most

women and their clinicians are unwilling to accept conservative management of detected

uterine polyps (240). In addition to hysteroscopic surgical interventions such as polypectomy

and myomectomy, other uterine sparing therapeutic interventions such as fitting of LNG-IUS

or endometrial ablation, may not be successfully completed. For the purposes of the model it

was assumed that all treatments were successfully performed given that the objective of the

economic analysis was to examine the cost-effectiveness of diagnostic testing based upon test

performance and accuracy, rather than an assessment of treatment efficacy.

Despite the need for assumptions arising in response to the ‘holes’ in the evidence base, the

analytic modelling methodology allowed an extensive, comprehensive, contemporary and

clinically relevant evaluation to be produced. An all-embracing randomised controlled trial,

assessing the wide range of diagnostic strategies defined in the models would be a huge,

impractical and ultimately futile undertaking.

257

Comparisons with existing guidance

Current NICE guidance published in 2007 for the management of HMB (142) recommends

that TVS should be the diagnostic modality of choice when testing is considered and OPH

employed if TVS is inconclusive. This advice was based upon a cost-effectiveness analysis,

limited to the assessment of three single testing options; TVS, SIS and OPH. The outcome

‘cost per correct diagnosis’ did not take into account the range of pathologies under

consideration for HMB, and the optimal treatments upon knowledge of the underlying

diagnosis. A fuller evaluation is necessary in order to consider the consequences of erroneous

diagnosis given that the raison d’etre of formulating a diagnosis is to inform and optimise

clinical management. Women may serendipitously receive appropriate treatment following a

false diagnosis, but more often than not, misdiagnosis results in misallocation of resources

and consequent morbidity attributable to unnecessary procedures. In contrast, the

comprehensive analysis performed in this thesis reflects the reality of diagnostic evaluation in

day to day clinical practice in the UK.

258

CHAPTER 11

CONCLUSIONS

Summary of the findings of the systematic reviews of outpatient

hysteroscopy

The systematic reviews of outpatient hysteroscopy showed that by making small adjustments

to the way the procedure is carried out can affect the pain experienced by the patient. The

reviews suggest that women attending for an outpatient hysteroscopy should take a non-

steroidal anti-inflammatory prior to arriving in clinic to try and reduce pain post-operatively.

When considering the effect on pain, the clinician performing the procedure may choose to

use a flexible hysteroscope, as this was shown to be less painful. However, there were more

failed procedures with flexible scopes and they are more expensive to maintain thus the

clinician may decide that the benefits of using a rigid hysteroscope outweigh the increased

level of pain, particularly when the pain scores were generally low overall (1.2-4.0). Although

there was no difference between carbon dioxide and normal saline in terms of pain scores,

normal saline has many other advantages (quicker procedure, reduced shoulder tip pain, acts

as a lavage to wash away blood and bubbles and it can be used during bipolar operative

procedures) that may prompt clinicians to favour it. Although injectable local anaesthetics

were shown to reduce pain, it was unclear how clinically significant this finding was,

especially when it was found that using the vaginoscopic technique reduces pain when

compared to the traditional procedure. The vaginoscopic technique should be the default

259

method for conducting outpatient hysteroscopy. Finally, the use of cervical preparation was

shown to not affect dilatation of the cervix or reduce pain during the procedure. Furthermore,

although the systematic review in this thesis found no significant increase in side effects with

the use of misoprostol, other meta-analyses of its use in inpatient procedures have found an

association with genital tract bleeding and gastrointestinal symptoms (24;106;107).These

findings led to the conclusion that routine cervical preparation with misoprostol prior to

outpatient hysteroscopy should be avoided.

An important consideration when interpreting the results of the systematic reviews is that the

primary studies tended to use hysteroscopes of 5.5mm in diameter which is much larger than

the outpatient hysteroscopes being used today (approximately 3mm). Thus the findings of the

reviews of local anaesthetic, cervical preparation and scope type may have even less clinical

significance than was suggested by the results.

Ultimately, using a small hysteroscope, via a vaginoscopic approach with normal saline is

likely to be acceptable to the majority of women who attend for outpatient hysteroscopy.

Local anaesthetic and cervical preparation should not routinely be used, however, future

research should investigate whether certain populations might benefit from these interventions

(postmenopausal or nulliparous women).

260

Summary of the comparison of the SIGN and GRADE methods for

assessing evidence quality

The comparison of the SIGN and GRADE methods for grading evidence quality in clinical

guidelines found that SIGN gives higher gradings of evidence quality than GRADE. This may

be because GRADE underestimates the quality but it seems more likely that this is due to

SIGN overestimating the quality due to its less rigorous assessment of methodology. GRADE

was found to be difficult to use because of the complex method of assessment and although

the intention is that GRADE is standardised, there are numerous opportunities for subjectivity

to be introduced. Neither GRADE nor SIGN appear to be the ideal method for classifying

and displaying evidence quality. A hybrid version which performs an in-depth assessment but

displays this data graphically rather than in tabulated form may result in more ‘user-friendly’

clinical guidance.

Summary of findings of the economic analysis of HMB

The economic analysis identified two potentially cost-effective testing strategies for the

investigation of women with HMB. These were initial testing with OPH alone or in

combination with EBx. To adopt a strategy of OPH, £360 needs to be invested to gain one

more woman satisfied at one year compared to a strategy of empirical treatment with a LNG-

IUS. Although a testing strategy of OPH +EBx is marginally more effective, the ICER is

approximately £21,000 to gain one more satisfied patient compared to OPH. This

approximates to around £26,500 per QALY which falls within the £20,000-£30,000 per

QALY threshold. These findings were stable during sensitivity analyses, varying model

261

inputs including disease prevalence, test feasibility and accuracy, with OPH remaining more

cost-effective than the LNG-IUS reference strategy even at relatively low willingness to pay

thresholds. In women wishing to preserve their fertility, therapy with endometrial ablation and

hysterectomy are contra-indicated. SIS was cost-effective in this situation, with an ICER of

approximately £2000, but for an additional financial outlay of £2720, testing instead with

OPH produces a further satisfied patient, which is likely to be considered affordable and

worthwhile by the NHS. Sensitivity analysis also showed SIS to be a cost-effective testing

option along with the combination testing strategy of OPH + EBx within the context of

traditional multi-stop pathways, although such service models are likely to diminish over time

with the on-going improvement and increasing availability of portable diagnostic health

technologies. Outpatient hysteroscopy remains the most cost-effective testing option if a

scenario is envisaged where only women with HMB refractory to the LNG-IUS,

recommended first line treatment in general practice (142), are referred to secondary care for

investigation. In this situation, it was estimated that £5480 extra funding is necessary for each

additional woman satisfied. No other testing strategies fell within plausible willingness to pay

ranges.

Therefore, the data are consistent in supporting OPH as the diagnostic testing strategy of

choice for women referred to secondary care with HMB irrespective of their desire for future

fertility and regardless of pre-referral treatment with the LNG-IUS. A combination strategy of

OPH+EBx may be cost-effective at the upper NICE willingness to pay threshold, in women

without a desire to retain their fertility who have not undergone pre-referral treatment with the

LNG-IUS or in women investigated through a traditional multi-visit pathway.

262

Summary of the role of outpatient hysteroscopy as investigated by this

thesis

Outpatient hysteroscopy is clearly an important intervention in gynaecological care not only

as a diagnostic test but also as an operative procedure (111;241). However, there was no

consensus on how to best perform the procedure until the systematic reviews and meta-

analyses in this thesis were produced. The results led to development of a clinical guideline

which is now available from the RCOG (website http://www.rcog.org.uk/womens-

health/clinical-guidance/hysteroscopy-best-practice-outpatient-green-top-59) and is

influencing patient care.

The economic analysis confirmed that outpatient hysteroscopy is not only cost-effective when

compared to other outpatient diagnostic test but that it is likely to be the most cost-effective

test for investigating women with HMB when used in a contemporary setting.

263

APPENDIX 1

A1.1 Search strategy for systematic review and meta-analysis of local

anaesthesia for pain control during outpatient hysteroscopy

1.1a Medline (1950- September 2008)

1. HYSTEROSCOPY/

2. Hysteroscopy.ti,ab

3. 1 OR 2

4. Vaginoscopy. Ti,ab

5. 3 OR 4

6. ANAESTHETICS, LOCAL/ OR LIDOCAINE/ OR BUPIVICAINE/ OR

ANAESTHESIA, LOCAL/ OR PROCAINE

7. “local anaesthetic”.ti,ab

8. (local AND anaesthe*).ti,ab

9. 6 OR 7 OR 8

10. 5 AND 9

1.1b EMBASE (1980 to September 2008)

1. HYSTEROSCOPY/


3. 1 OR 2

4. VAGINOSCOPY/

5. Vaginoscopy.ti,ab

6. 4 OR 5

7. 3 OR 6

8. LOCAL ANESTHTEIC AGENT/ OR LOCAL ANESTHESIA/ OR

BUPIVICAINE/ OR LIDOCAINE/ OR MEPIVICAINE/ OR ROPIVICAINE/ OR

PRILOCAINE/ OR LEVOBUPIVICAINE/ OR TETRACAINE/


10. 8 OR 9

11. 7 AND 10

264

1.1c CINAHL (1981 to September 2008)

1. HYSTEROSCOPY/



4. 1 OR 2 OR 3

5. ANESTHESIA, LOCAL/ OR ANESTHTICS, LOCAL/


7. 5 OR 6

8. 4 AND 7

1.1d The Cochrane Library (Cochrane Central Register of Controlled Trials)

1. hysteroscopy AND anaesthetic

265

A1.2 Search strategy for the systematic review of analgesia for pain control



1. HYSTEROSCOPY/


3. 1 OR 2


5. 3 OR 4

6. ANALGEISA/

7. ANALGESICS/ OR ANALGESICS, NON-NARCOTIC/ OR ANALGESICS,

OPIOID/

8. Analges*.ti,ab

9. 6 OR 7 OR 8

10. 5 AND 9


1. HYSTEROSCOPY/


3. 1 OR 2

4. VAGINOSCOPY/


6. 4 OR 5

7. 3 OR 6

8. ANALGESIA/

9. ANALGESIC AGENT/

10. Analges*.ti,ab

11. 8 OR 9 OR 10

12. 7 AND 11

266


1. HYSTEROSCOPY/



4. 1 OR 2 OR 3

5. ANALGESIA/

6. analgesia.ti,ab

7. analges*.ti,ab

8. ANALGESICS/ OR ANALGESICS, OPIOID/ OR ANALGESICS,

NONNARCOTIC/ OR NARCOTICS/ OR TRAMADOL/ OR

ANTIINFLAMMATORY AGENTS, NON-STEROIDAL/

9. 5 OR 6 OR 7 OR 8

10. 4 AND 9


1. hysteroscopy AND analgesia

267

A1.3 Search strategy for the systematic review of conscious sedation for



1. HYSTEROSCOPY/


3. 1 OR 2


5. 3 OR 4

6. CONSCIOUS SEDATION/

7. “conscious sedation”.ti,ab

8. 6 OR 7

9. HYPNOTICS AND SEDATIVES/

10. Sedative.ti,ab

11. 9 OR 10

12. 8 OR 11

13. 5 AND 12


1. HYSTEROSCOPY/


3. 1 OR 2

4. VAGINOSCOPY/


6. 4 OR 5

7. 3 OR 6



10. SEDATIVE AGENT/

11. Sedative.ti,ab

12. 8 OR 9 OR 10 OR 11

13. 7 AND 12

268


1. HYSTEROSCOPY/



4. 1 OR 2 OR 3



7. HYPNOTICS AND SEDATIVES/ OR SEDATIVES, BARBITURATE/ OR

SEDATIVES, NONBARBITURATE/

8. Sedate*.ti,ab

9. 5 OR 6 OR 7 OR 8

10. 4 AND 9


1. hysteroscopy AND sedation

269

A1.4 Search strategy for the systematic review of cervical preparation for


1.4a Medline (1950- February 2010)

1. HYSTEROSCOPY/


3. 1 OR 2


5. 3 OR 4

6. MISOPROSTOL/ OR DINOPROSTONE

7. LAMINARIA/

8. PROSTAGLANDINS/ OR PROSTAGLANDINS, SYNTHETIC/

9. ESTROGENS/

10. (oestrogen OR estrogen).ti,ab

11. PROGESTERONE/

12. PROGESTINS/

13. CERVICAL RIPENING/

14. "cervical preparation".ti,ab

15. laminaria.ti,ab

16. prostaglan*.ti,ab

17. progest*.ti,ab

18. "cervical ripening".ti,ab

19. 6 AND 7 AND 8 AND 9 AND 10 AND 11 AND 12 AND 13 AND 14 AND 15

AND 16 AND 17 AND 18

20. 5 AND 20

270

1.4b EMBASE (1980 to February 2010)

1. HYSTEROSCOPY/


3. 1 OR 2

4. VAGINOSCOPY/


6. 4 OR 5

7. 3 OR 6

8. MISOPROSTOL/ OR GEMEPROST/ OR PROSTAGLANDIN E2/ OR

MIFEPRISTONE/ OR DILAPAN/ OR PROSTAGLANDIN/ OR UTERINE

CERVIX DILATATION/

9. UTERINE CERVIX RIPENING/

10. 8 OR 9


12. 10 OR 11

13. LAMINARIA/

14. laminaria.ti,ab

15. 13 OR 14

16. 12 OR 15

17. 7 AND 16

1.4c CINAHL (1981 to February 2010)

1. HYSTEROSCOPY/



4. 1 OR 2 OR 3

5. MISOPROSTOL.ti,ab

6. Laminaria.ti,ab

7. prostaglandins.ti,ab

8. (oestrogen OR estrogen).ti,ab

9. progest*.ti,ab


11. "cervical prep*".ti,ab

12. MISOPROSTOL/

13. PROSTAGLANDINS/

14. ESTROGENS/

15. PROGESTERONE/

16. CERVIX DILATATION AND EFFACEMENT/

17. 5 OR 6 OR 7 OR 8 OR 9 OR 10 OR 11 OR 12 OR 13 OR 14 OR 15 OR 16

18. 4 AND 17


1. hysteroscopy AND cervical

271

A1.5 Search strategy for the systematic review of the effect of the

vaginoscopic approach to outpatient hysteroscopy on pain


1. HYSTEROSCOPY/ae,mt [ae=Adverse Effects, mt=Methods]

2. hysteroscopy.ti,ab


4. “no touch”.ti,ab

5. HYSTEROSCOPY/

6. Vaginoscop*.ti,ab

7. 2 OR 5

8. 1 OR 3 OR 4 OR 6

9. 7 AND 8


1. exp HYSTEROSCOPY/

2. hysteroscopy.ti,ab

3. 1 OR 2

4. VAGINOSCOPY/



7. 1 OR 2

8. 3 OR 4 OR 5


10. 7 OR 9

11. 6 AND 10

272



2. HYSTEROSCOPY/AE,MT,AE,MT [AE=Adverse Effects, MT=Methods,

AE=Adverse Effects, MT=Methods]



5. 2 OR 3 OR 4

6. 1 AND 5


8. 5 OR 7

9. HYSTEROSCOPY/

10. 1 OR 9

11. 8 AND 10


1. hysteroscopy AND (vaginoscopy OR vaginoscopic OR "no-touch")

273

A1.6 Search strategy for the systematic review of the effect of distension

media on pain during outpatient hysteroscopy


1. HYSTEROSCOPY/


3. 1 OR 2


5. 3 OR 4

6. DEXTRANS/ OR SODIUM CHLORIDE/ OR MANNITOL/ OR SORBITOL/

7. “distension media”.ti,ab

8. (uter* AND disten*).ti,ab

9. CARBON DIOXIDE

10. “carbon dioxide”.ti,ab

11. 6 OR 7 OR 8 OR 9 OR 10

12. 5 AND 11


1. HYSTEROSCOPY/


3. 1 OR 2

4. VAGINOSCOPY/


6. 4 OR 5

7. 3 OR 6



10. SODIUM CHLORIDE/ OR CARBON DIOXIDE/

11. 8 OR 9 OR 10

12. 7 AND 11

274


1. HYSTEROSCOPY/



4. 1 OR 2 OR 3


6. NORMAL SALINE/ OR SALINE SOLUTION, HYPERTONIC/ OR SODIUM

CHLORIDE/

7. DEXTRANS/

8. Dextrans.ti,ab

9. MANNITOL/

10. Mannitol.ti,ab

11. CARBON DIOXIDE/

12. “carbon dioxide”.ti,ab


14. 5 OR 6 OR 7 OR 8 OR 9 OR 10 OR 11 OR 12 OR 13 Or 14 OR 15

15. 4 AND 14


1. hysteroscopy AND distension

275

A1.7 Search strategy for systematic review of the effect on pain of the type

of hysteroscope used for outpatient hysteroscopy


1. HYSTEROSCOPY/


3. vaginoscopy. Ti,ab

4. flexible.ti,ab

5. flex*.ti,ab

6. rigid.ti,ab

7. rigid*.ti,ab

8. 1 OR 2 OR 3

9. 4 OR 5 OR 6 OR 7

10. 8 AND 9


1. HYSTEROSCOPY/


3. VAGINOSCOPY/


5. Flex*.ti,ab

6. Rigid.to,ab

7. Rigid*.ti,ab

8. Flexible.ti,ab

9. 5 OR 6 OR 7 OR 8

10. 1 OR 2 OR 3 OR 4

11. 9 AND 10

276


1. HYSTEROSCOPY/



4. Flexible.ti,ab

5. Flex*.ti,ab

6. Rigid.ti,ab

7. Rigid*.ti,ab

8. 4 OR 5 OR 6 OR 7

9. 1 OR 2 OR 3

10. 8 AND 9


1. hysteroscopy AND (flexible OR rigid)

277

APPENDIX 2

Data retrieval form for systematic review and meta-analysis of local

anaesthesia for pain control during outpatient hysteroscopy

The following form was used for data extraction from studies selected for use in the

systematic review of local anaesthesia. The data extraction forms used for the other

systematic reviews were adapted to evaluate the different interventions by changing the

selection criteria and the secondary outcomes.

Reviewer ……………. Paper No. ……….. Name of 1st author & year…………

Ring the appropriate category(ies)

i) population – women undergoing out-patient hysteroscopy yes / no

ii) intervention – use of local anaesthetic yes / no

iii) outcome – patient experience i.e. pain / feasibility yes / no

Select this study (i-iii inclusive) yes / no

if no reject & specify reason

……………………………………………………………………………………………..

Selection criteria (therapy study):

278

Study Design:

(1) Study Design RCT Yes [ ] No [ ]

(2) Was the study described as randomized (this includes the use of words such as

randomly, random, and randomization)?

Yes [ ] No [ ]

Was the method to generate the sequence of randomization described?

Yes [ ] No [ ]

If yes was it:

appropriate (table of random numbers, computer generated etc)? [ ]

or

Inappropriate (patients were allocated alternately, or according to DOB etc.) [ ]

(3) Was the study described as blinded? Double [ ] Single [ ] Not blinded [ ] NR [ ]

If double blind was the method:

Appropriate (identical placebo, active placebo, dummy, etc.). [ ]

Or

Inappropriate (e.g. comparison of pill v. injection without double dummy). [ ]

(4) Was there a description of withdrawals and dropouts? Yes [ ] No [ ]

(Participants who were included in the study but did not complete the observation period or

who were not included in the analysis must be described. The number and the reasons for

withdrawal in each group must be stated. If there were no withdrawals, it should be stated in

the article. If there is no statement on withdrawals, this item must be given no points.)

Population:

(1) Type of patient Premenopausal [ ] Postmenopausal [ ] Both [ ]

Data Retrieval:

279

(2) No. patients recruited A original population n=……

B Pre-enrolment /randomisation n=……

exclusions (reasons e.g. pop characteristics)

…………………………………………………….

C actually recruited (A-B) n=……

D post-enrolment /randomisation n=……

exclusions (reasons e.g. missing data, drop-outs)

…………………….………………………………..…

E analysable data (C-D) n=……

(3) Type of procedure Diagnostic [ ] Operative [ ]

Speculum used Yes [ ] No [ ] Sometimes [ ]

Scope type and size......................................................................................................

Distension medium used ......................................……………….............................

Pipelle sample taken Yes [ ] No [ ] Sometimes [ ]

Intervention:

(1) Type & nature of anaesthesia

No in intervention group……………………………………………….....…….......……….

What anaesthetic?

…………………………………….…Dose……..…………….....…….……....…………

Method of administration (e.g. Direct cervical / paracervical)……………………………

When administered?……………..……………………….……………………………….

No. in control group ……………………………….…………………….………………..

Control method (e.g. placebo used etc.)......……………………………………………..

280

(2) Description of technique (i.e. could we reproduce it) Adequate / Inadequate

Outcome:

(1) Method of assessing pain e.g. pain scores, VAS, use of post op analgesia, BP

measurements, cortisol

…………………………………………………………………………………………………

…………………………………………………………………………………………………

(2) Was pain relief / assessment defined? yes / no

If so, how.......................................................................................................................................

(3) Pain results

(4) Completeness of Follow up (%) >90 / 81-90 / <81 %

Secondary outcomes:

(1) Fail to perform procedure NR inter grp. n=……control grp n= .................

(2) Need for post-op analgesia NR inter grp. n=….… control grp. n=….........

(3) Complication rate NR inter grp. n=….… control grp. n=….........

(4) Vasovagal reactions NR inter grp. n=….… control grp. n=….........

Has ‘vasovagal’ been defined? yes / no

If so, how?.........................................................................................................................

...................................................................................................................................................

Intervention group Control group

Number in group

Values (e.g. mean, sd)

Pain scores

281

(5) Patient admitted NR inter grp. n=….… control grp. n=….........

(6) Other (e.g. patient satisfaction) NR inter grp. n=….… control grp. n=….........

2x2 Contingency tables (RCT / controlled studies only):

Local Anaesthetic

Outcome

Intervention

Pain No / acceptable pain Total

Anaesthetic

Control / Comparison

Total

Vasovagal episodes

Outcome

Intervention

Outcome present Outcome absent Total

Surgery


Total

Other

Outcome

Intervention

Outcome present Outcome absent Total

Surgery


Total

282

APPENDIX 3

Jadad method for scoring the Quality of Randomised Controlled Trials (50)

1. Was the study described as randomized (this includes the use of words such as

randomly, random, and randomisation)?

2. Was the study described as double blind?

3. Was there a description of withdrawals and dropouts?

Scoring the items:

Either give a score of 1 point for each ”yes” or 0 points for each “no.” There are no in-

between marks.

Give 1 additional point if:

For question 1, the method to generate the sequence of

randomisation was described and it was appropriate

(table of random numbers, computer generated, etc.).

and/or:

If for question 2 the method of double blinding was

described and it was appropriate (identical placebo,

active placebo, dummy, etc.).

Deduct 1 point if: For question 1, the method to generate the sequence of

randomization was described and it was inappropriate

(patients were allocated alternately, or according to date

of birth, hospital number, etc.)

and/or:

For question 2, the study was described as double blind

but the method of blinding was inappropriate (e.g.,

comparison of tablet vs. injection with no double

dummy).

Guidelines for Assessment

1. Randomisation

A method to generate the sequence of randomisation will be regarded as appropriate if it

allowed each study participant to have the same chance of receiving each intervention and

the investigators could not predict which treatment was next. Methods of allocation using

date of birth, date of admission, hospital numbers, or alternation should be not regarded as

appropriate.

283

2. Double blinding

A study must be regarded as double blind if the word “double blind” is used. The method

will be regarded as appropriate if it is stated that neither the person doing the assessments

nor the study participant could identify the intervention being assessed, or if in the

absence of such a statement the use of active placebos, identical placebos, or dummies is

mentioned.

3. Withdrawals and dropouts

Participants who were included in the study but did not complete the observation period or

who were not included in the analysis must be described. The number and the reasons for

withdrawal in each group must be stated. If there were no withdrawals, it should be stated

in the article. If there is no statement on withdrawals, this item must be given no points.

284

APPENDIX 4

‘Best Practice in Outpatient Hysteroscopy’ Guideline

Reproduced from: Royal College of Obstetricians and Gynaecologists. Best Practice in

Outpatient Hysteroscopy. Green-top Guideline No. 59. London: RCOG; 2011, with the

permission of the Royal College of Obstetricians and Gynaecologists.

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

APPENDIX 5

GRADE tables for ‘Best Practice in Hysteroscopy’ Guideline recommendations with corresponding SIGN

gradings

A5.1 Analgesia

Do analgesics given prior to diagnostic hysteroscopy reduce the pain felt during the procedure?

Routine use of opiate analgesia prior to outpatient hysteroscopy should be avoided as it may cause adverse side effects

Question: Should tramadol be used for pain relief prior to outpatient hysteroscopy?


Quality Importance

No of studies

Design Risk of



Tramadol Control Relative

(95% CI)

Absolute

Pain with im tramadol (measured with: Visual analogue scale; range of scores: 0-20; Better indicated by lower values)

1 randomised trials

very serious

1




3 40 40 - MD 3.3 lower (4.96 to 1.64

lower) VERY LOW

CRITICAL

Pain with iv tramadol (measured with: Visual analogue scale; range of scores: 0-10; Better indicated by lower values)

1 randomised trials

serious4 no serious

inconsistency no serious indirectness


3 25 25 -

VERY LOW

CRITICAL

Side effects with iv tramadol

1 randomised trials

serious5 no serious



3 4/25

(16%) 2/25 (8%)

- 80 fewer per 1000 (from 80 fewer to 80 fewer)

VERY LOW

CRITICAL

306

1 Randomisation not described and no blinding. Unclear who assessed the pain scores but likely to be the surgeons.

2 Only one small study.

3 This is the only study identified.

4 The study does not accurately report the point estimate and confidence intervals it displays them on a graph that cannot be reliably read.

Reports that the result is significant p<0.012 and p<0.008. 5 Not clear whether the study was blinded.

Question: Should sublingual buprenorphine be used for pain relief prior to outpatient hysteroscopy?


Quality Importance

No of studies

Design Risk of



Sublingual buprenorphine

Control Relative (95% CI)

Absolute

Pain (measured with: Visual analogue scale; range of scores: 0-10; Better indicated by lower values)

1 randomised trials

serious1 no serious

inconsistency serious

2 serious

3 reporting bias

4 80 84 - MD 0.1 higher

(0.27 lower to 0.47 higher)

VERY LOW

CRITICAL

Presence of side effects associated with buprenorphine

1 randomised trials

serious1 no serious


very serious

3,5


(38.8%) 0% OR 107.55

(6.44 to 1796.46)

- VERY LOW

IMPORTANT

1 Not clear whether the study was blinded.

2 Pain was assessed by a nurse who was observing the facial expressions of the patients and not the patients themselves.

3 This is single study with a small population.

4 This was the only study identified.

5 The confidence interval is very wide.

SIGN grading: B

307

Patients, without contra-indications, should be advised to consider taking standard doses of non-steroidal anti-inflammatory agents

around one hour before their scheduled outpatient hysteroscopy appointment with the aim of reducing pain in the immediate post-

operative period.

Question: Should non-steroidal anti-inflammatory drugs (NSAIDs) be used for pain relief prior to outpatient hysteroscopy?


Quality Importance

No of studies

Design Risk of bias Inconsistency Indirectness Imprecision Other

considerations NSAIDs Control

Relative (95% CI)

Absolute

Visual analogue scale pain score with oral diclofenac (range of scores: 0-10; Better indicated by lower values)

1 randomised trials

no serious risk of bias



serious reporting bias 92 89 - MD 0 higher (0.79 lower to 0.79 higher)

LOW

CRITICAL

Visual analogue scale pain score with im ketorolac (range of scores: 0-10; Better indicated by lower values)

1 randomised trials

serious1 no serious



3 12 12 - MD 1.1 lower (3.35 lower

to 1.16 higher) VERY LOW

CRITICAL

Visual analogue scale pain score with oral mefenamic acid (range of scores: 0-10; Better indicated by lower values)

1 randomised trials





5 49 46 - MD 0 higher (0 to 0

higher) LOW

CRITICAL

1 There is no description of randomisation and the study is not blinded.

2 This is a single study with a very small population.

3 This was the only study identified.

4 This is only one small study.

5 This is the only identified study.

SIGN grading: B

308

A5.2 Cervical Preparation

Does cervical preparation reduce uterine trauma, failure to access the uterine cavity or pain associated with outpatient

hysteroscopy?

Cervical preparation prior to outpatient hysteroscopy should not be used in the absence of any evidence of benefit in terms of reduction

in pain, rates of failure or uterine trauma.

Question: Should Vaginal Prostaglandins be used in women undergoing outpatient hysteroscopy?


Quality Importance

No of studies

Design Risk of



Vaginal Prostaglandins


Absolute

Premenopausal women: pain score during dilatation (range of scores: 0-10; Better indicated by lower values)

1 randomised trials

very serious

1



very serious2 reporting bias

3 22 21 - 0 higher (0 to 0

higher)

VERY LOW

CRITICAL

Premenopausal: pain reduction during hysteroscopy

1 randomised trials

serious4 no serious


very serious5 reporting bias

3 6/13

(46.2%) 12/23

(52.2%) RR 0.88 (0.44 to

1.79)

63 fewer per 1000 (from 292 more to

412 more)

VERY LOW

CRITICAL

Postmenopausal: pain reduction during the procedure (measured with: visual analogue scale; Better indicated by lower values)

1 randomised trials

serious6 serious

7 no serious


reporting bias3 60 60 - 0 higher (0 to 0

higher)

VERY LOW

CRITICAL

Mixed population: pain (measured with: Visual analogue scale; range of scores: 0-10; Better indicated by lower values)

1 randomised trials

very serious

8




3 50 50 - 0 higher (0 to 0

higher)

VERY LOW

CRITICAL

Mixed population: pain after dilatation (measured with: Visual analogue scale; range of scores: 0-100; Better indicated by lower values)

1 randomised trials




serious9,10

reporting bias3 50 51 - MD 12.3 lower

(13.69 to 10.91 lower)

LOW

CRITICAL

Premenopausal: adequate dilatation

1 randomised trials

very serious

11





(90.9%) 6/21

(28.6%) OR 25 (4.41 to 141.68)

623 more per 1000 (from 352 more to

697 more)

VERY LOW

IMPORTANT

309

Postmenopausal- need for dilatation

1 randomised trials

serious12



serious13


(17.2%) 12/59

(20.3%) OR 0.82 (0.32 to

2.06)

30 fewer per 1000 (from 128 fewer to

141 more)

VERY LOW

IMPORTANT

Needs dilatation- mixed population

1 randomised trials

very serious

8




3 15/50

(30%) 11/50 (22%)

OR 1.52 (0.62 to

3.74)

80 more per 1000 (from 71 fewer to

293 more)

VERY LOW

IMPORTANT

Force needed to dilate at 6mm- mixed population (Better indicated by lower values)

1 randomised trials





reporting bias3 50 51 - SMD 0.48 lower

(0.88 to 0.09 lower)

MODERATE

IMPORTANT

Misoprostol in postmenopausal women- failed hysteroscopies

1 randomised trials

serious12



serious14


(6.7%) 4/60

(6.7%) OR 1 (0.24

to 4.2) 0 fewer per 1000 (from 50 fewer to

164 more)

VERY LOW

IMPORTANT

Misoprostol in premenopausal women- failed hysteroscopies

1 randomised trials




very serious15


(35%) 1/24

(4.2%) OR 12.38 (1.37 to 112.1)

308 more per 1000 (from 15 more to

788 more)

VERY LOW

IMPORTANT

Cervical lacerations

3 randomised trials




serious14

reporting bias16

8/125 (6.4%)

13/131 (9.9%)

OR 0.59 (0.22 to

1.55)

38 fewer per 1000 (from 76 fewer to 47

more)

LOW

CRITICAL

Cervical Bleeding

4 randomised trials


serious17


serious13

reporting bias18

21/135 (15.6%)

12/141 (8.5%)

OR 1.32 (0.52 to 3.4)

24 more per 1000 (from 39 fewer to

155 more)

VERY LOW

CRITICAL

1 Randomisation not described and not blinded.

2 Only one study with a very small population and a large effect so results are less likely to be accurate.

3 Only a single study was identified.

4 33% of the women randomised to misoprostol did not undergo hysteroscopy and therefore could not be assessed.

5 The population size is very small. The confidence interval includes 'no effect'.

6 Probable reporting bias as multiple evaluations of pain are reported. Blinding suggested but unclear.

7 Results are confusing with no clear evidence of effect.

8 No blinding as compared to nothing rather than placebo.

9 Only one study with a small population so results less likely to be accurate.

10 Single study with population size that was calculated based on the outcome 'dilatation' and not pain.

11 Randomisation not described and study not blinded.

12 Blinding suggested but unclear.

310

13 The confidence interval contains no effect.

14 Confidence interval includes 'no effect'. Also a small number of events.

15 A single study with a small population and wide confidence intervals.

16 Only three small trials.

17 Significant heterogeneity which can only be partially explained by different administration times and doses.

18 Only four small trials.

Question: Should Mifepristone be used in women undergoing outpatient hysteroscopy?


Quality Importance

No of studies

Design Risk of bias Inconsistency Indirectness Imprecision Other

considerations Mifepristone Control

Relative (95% CI)

Absolute

Pain after cervical dilatation (measured with: Visual analogue scale; range of scores: 0-100; Better indicated by lower values)

1 randomised trials





2 28 30 - MD 3.6 lower (17.42

lower to 10.22 higher) LOW

CRITICAL

1 Results are from one small study with a wide confidence interval.

2 Only one study identified.

SIGN grading: A

311

A5.3 Type of Hysteroscope

Should rigid or flexible hysteroscopes be used routinely in the outpatient setting?

Flexible hysteroscopes are associated with less pain during outpatient hysteroscopy compared with rigid hysteroscopes. However, rigid

hysteroscopes may provide better images, fewer failed procedures, quicker examination time and reduced cost Thus there is insufficient

evidence to recommend preferential use of rigid or flexible hysteroscopes for diagnostic, outpatient procedures and choice of

hysteroscope should be left to the discretion of the operator

Question: What type of hysteroscope should be used for outpatient hysteroscopy?


Quality Importance

No of studies

Design Risk of



Flexible hysteroscopes

Rigid hysteroscopes

Relative (95% CI)

Absolute

Pain score (measured with: Visual analogue scale; range of scores: 0-10; Better indicated by lower values)

2 randomised trials


1 no serious inconsistency



3 110 115 - not

pooled LOW

IMPORTANT

Failures (assessed with: Occurrence vs. non-occurrence)

2 randomised trials


1 no serious inconsistency


serious2,4


(4.5%) 0/115 (0%)

OR 11.81 (0.64 to 217.75)

- LOW

IMPORTANT

1 The only limitation is that the studies were not double blinded however this would be impossible to do.

2 Small sample sizes encourage imprecision.

3 Only two small studies. Likely that studies finding no advantage of one hysteroscope over the other are not published.

4 Small number of events.

SIGN grading: B

312

A5.4 Distension Media

Which uterine distension medium should be used during outpatient hysteroscopy?

The choice of distension media, for routine outpatient hysteroscopy, between carbon dioxide and normal saline should be left to the

discretion of the operator as neither is superior in reducing pain, although uterine distension with normal saline appears to reduce the

incidence of vasovagal episodes.

Uterine distension with Normal saline allows outpatient diagnostic hysteroscopy to be completed more quickly compared with carbon

dioxide.

Question: Which is the best distension medium for outpatient hysteroscopy?


Quality Importance

No of studies

Design Risk of



Normal saline

Carbon dioxide

Relative (95% CI)

Absolute


6 randomised trials

very serious

1

serious2 no serious


none 540 569 - SMD 0.34 higher (0.12 lower to 0.8 higher)

VERY LOW

CRITICAL

Shoulder tip pain

5 randomised trials

very serious

1




none 9/472 (1.9%)

55/448 (12.3%)

OR 0.19 (0.09 to 0.4)


fewer)

LOW

CRITICAL

Vasovagal episodes

3 randomised trials

very serious

1





(2.4%) 18/210 (8.6%)

OR 0.31 (0.12 to 0.82)


fewer)

VERY LOW

IMPORTANT

Procedure time (Better indicated by lower values)

4 randomised trials

very serious

1


4



none 434 401 - SMD 1.32 lower (1.48 to 1.17 lower)

LOW

IMPORTANT

Unsatisfactory hysteroscopic view

1 randomised trials

very serious

1




6 4/100

(4%) 19/100 (19%)

RR 4.75 (1.61 to 16.4)

712 more per 1000 (from 116 more to 1000

more)

VERY LOW

CRITICAL

313

1 Poor randomisation methods (quasi-randomised in some cases) and lack of allocation concealment in most studies. Neither patient nor

operator was blinded but this would not be possible. 2 High heterogeneity that cannot be explained by differences in technique.

3 Outcome reported in just three of the six identified trials.

4 High heterogeneity but all studies in favour of normal saline and heterogeneity can be explained by the two studies that don't using a

speculum showing the largest effect size. 5 Single study therefore imprecise.

6 This is the only study that reports a difference.

SIGN grading: A

314

A5.5 Local anaesthesia and cervical dilatation

Should topical local anaesthetic be administered prior to outpatient hysteroscopy?

Injection of local anaesthetic into the cervical canal does not reduce pain during diagnostic outpatient hysteroscopy

Should injectable local anaesthetic be administered to the cervix and / or para-cervix prior to outpatient hysteroscopy?

Injection of local anaesthetic into or around the cervix is associated with reduction in pain experienced during outpatient diagnostic

hysteroscopy. However, it is unclear how clinically significant this reduction in pain is. Consideration should be given to the routine

administration of intracervical or paracervical local anaesthetic, particularly in postmenopausal women.

Routine administration of intracervical or paracervical local anaesthetic is not indicated to reduce the incidence of vasovagal reaction.

Question: Should local anaesthetic be used in women undergoing outpatient hysteroscopy?


Quality Importance

No of studies

Design Risk of



Local anaesthetic


Absolute

pain score with topical local anaesthetic (Better indicated by lower values)

2 randomised trials



1



3 312 306 - SMD 0.32 lower

(0.97 lower to 0.33 higher)

LOW

CRITICAL

pain score with transcervical anaesthetic (measured with: Visual analogue scale; range of scores: 0-10; Better indicated by lower values)

5 randomised trials

serious4 no serious



none 276 270 - SMD 0.11 lower (0.31 lower to 0.1

higher)

MODERATE

CRITICAL

Pain score with intracervical anaesthetic block (Better indicated by lower values)

3 randomised trials

serious5 no serious



7 123 116 - SMD 0.36 lower

(0.61 to 0.1 lower)

VERY LOW

CRITICAL

315

pain with paracervical anaesthetic block (Better indicated by lower values)

5 randomised trials

very serious

8

serious9 no serious

indirectness serious

10 none 335 337 - SMD 1.28 lower

(2.22 to 0.35 lower)

VERY LOW

CRITICAL

Vasovagal attacks

9 randomised trials

serious8 serious

11 no serious


none 40/635 (6.3%)

56/629 (8.9%)

OR 0.69 (0.45 to 1.05)


more)

LOW

IMPORTANT

1 Studies use different strengths of lignocaine which may explain the heterogeneity. However, stratifying for strength of anaesthetic agent

would result in multiple, small sub-groups. 2 Wide confidence intervals with a small effect size.

3 Only two studies included.

4 Poor allocation concealment.

5 Details of randomisation and allocation concealment are unclear.

6 Small population size.

7 Only 3 small studies identified.

8 Poor allocation concealment. Details of randomisation unclear and some studies not blinded.

9 Heterogeneity is high. When the study with greatest effect size is removed I2 is still 90%. Two studies do not overlap with any of the

other studies. 10

Wide confidence intervals. 11

Heterogeneous. The studies use different methods and use different administration of local anaesthetic.

SIGN grading: A

316

A5.6 Conscious sedation

Should conscious sedation be used to reduce pain associated with outpatient hysteroscopic procedures?

Conscious sedation should not be routinely used in outpatient hysteroscopic procedures, as it confers no advantage in terms of pain

control and patient satisfaction over local anaesthesia.

Question: Does conscious sedation reduce pain during outpatient hysteroscopy?


Quality Importance

No of studies

Design Risk of



Conscious sedation versus intracervical

block Control

Relative (95% CI)

Absolute


1 randomised trials

serious1 no serious



3 84 82 - MD 0.2 lower (0.43

lower to 0.03 higher)

VERY LOW

CRITICAL

1 Allocation concealment not clear. Not blinded.

2 Single study and small population therefore imprecise.

3 Only a single study identified.

SIGN grading: A

317

A5.7 Vaginoscopy

Does a vaginoscopic approach to outpatient hysteroscopy reduce pain and increase the feasibility of the procedure?

Vaginoscopy reduces pain during diagnostic rigid outpatient hysteroscopy.

Question: Should the vaginoscopic technique be used for outpatient hysteroscopy?


Quality Importance No of

studies Design

Risk of bias

Inconsistency Indirectness Imprecision Other

considerations Vaginoscopy Control

Relative (95% CI)

Absolute


4 randomised trials

serious1 no serious

inconsistency2



reporting bias3 472 439 - SMD 0.44 lower (0.65

to 0.22 lower) LOW

CRITICAL

Failed procedures

5 randomised trials

serious1 no serious



3 39/584

(6.7%) 25/550 (4.5%)

OR 1.28 (0.74 to 2.24)

12 more per 1000 (from 11 fewer to 51

more)

VERY LOW

IMPORTANT

1 Poor allocation concealment.

2 Heterogeneity can be explained by the fact that the traditional technique varied between studies.

3 Only a small number of trials and they mainly have positive results.

4 Wide confidence interval.

SIGN grading: A

318

APPENDIX 6

Decision trees for the economic analysis of heavy menstrual bleeding

The heavy menstrual bleeding decision trees are too large to be displayed in their entirety therefore an expanded branch from the outpatient

hysteroscopy tree is shown below. Table 3.2 shows the diagnosis made and the treatments given for all branches within the decision tree.

A6.1 Expanded branch from the outpatient hysteroscopy strategy for investigation of HMB

Satisfied

pSatisfied_Removal

Satisfied

pSatisfied_LngIUS_DUB

Not satisfied

#

LNG-IUSNot satisfied

#

RemoveShows polyp /SMF

pTPR_PolypsorSMF_OPH

Satisfied

pSatisfied_LngIUS_PolypSMF

Satisfied

pSatisfied_Removal

Not satisfied

#

Remove

EA but not done as

hysteroscopy shows

polyp/SMFNot satisfied

#

LNG-IUSShows polyp/SMFConfirmatory testShows endometrial

disease

pFPR_ED_OPH

Satisfied

pSatisfied_LngIUS_PolypSMF

Satisfied

pSatisfied_Removal

Not satisfied

#

Remove

EA but not done as

hysteroscopy shows

polyp/SMFNot satisfied

#

LNG-IUSNo intracavity

pathology/ DUB

#

Doesn't show

polyp /SMF

#

Succeed

pSuccessful_OPH

Satisfied

pSatisfied_Removal

Satisfied

pSatisfied_LngIUS_DUB

Not satisfied

#

LNG-IUSNot satisfied

#

Remove Shows polyp / SMFHyst D+C Fail

#

OPHResectable intrauterine

pathology (polyp/SMF)

pPolyporSMF

Fibroids <12/40

pFibroidSmall [+]

Fibroids >12/40

pFibroidLarge [+] Endometrial disease

pEndometrialDisease [+] Dysfunctional

uterine bleeding

# [+]

OPH alone

319

A6.2 Diagnosis and treatments from the decision tree for investigation of women with HMB

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2

Diagnosis from test 3 First Treatment Second treatment

LNG-IUS only Polyp / SMF LNG-IUS Resection- lesion

diagnosed at pre-

ablation

hysteroscopy

Fibroids <12 LNG-IUS EA

Fibroids >12 LNG-IUS EA

Hysterectomy if

bimanual exam

suggests >12 weeks size

GP appt

ED LNG-IUS Hysterectomy

DUB LNG-IUS EA

Hysterectomy

only

Polyp / SMF Hysterectomy GP appt

Fibroids <12 Hysterectomy GP appt

Fibroids >12 Hysterectomy GP appt

ED Hysterectomy GP appt

DUB Hysterectomy GP appt

OPH alone Polyp / SMF Polyp / SMF Resection LNG-IUS

Endometrial

disease

Polyp / SMF LNG-IUS Resection- lesion

diagnosed at pre-

ablation

hysteroscopy

DUB LNG-IUS Resection- lesion

diagnosed at pre-

ablation

hysteroscopy

Fibroids < 12 Normal cavity LNG-IUS EA

320

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


Polyp / SMF Resection but as normal

endometrium LNG-IUS

inserted

EA

ED Normal

endometrium

LNG-IUS EA

Fibroids >12 Normal cavity >12 by

bimanual exam

Hysterectomy GP appt

<12 by

bimanual exam

LNG-IUS

Polyp / SMF >12 by

bimanual exam


<12 by

bimanual exam

Resection but as normal

endometrium LNG-IUS

inserted

EA

ED >12 by

bimanual exam

Normal endometrium Hysterectomy GP appt

<12 by

bimanual exam

Normal endometrium LNG-IUS EA

ED ED Hyperplasia LNG-IUS Hysterectomy

Cancer Hysterectomy GP appt

Polyp / SMF Resection- histology

shows hyperplasia /

cancer so get LNG-IUS/

hysterectomy

Hysterectomy /

GP visit

DUB LNG-IUS Hysterectomy as

EBx prior to EA

shows ED

DUB DUB LNG-IUS EA


endometrium LNG-IUS

inserted

EA

321

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


TVS alone

and

SIS alone

Polyp / SMF Polyp / SMF Resection LNG-IUS

Fibroids <12 LNG-IUS Resection- lesion

diagnosed at pre-

ablation

hysteroscopy

ED Polyp / SMF Resection LNG-IUS


diagnosed at pre-

ablation

hysteroscopy

Fibroids <12 Fibroids < 12 LNG-IUS EA


endometrium LNG-IUS

inserted

EA

ED Normal

endometrium

LNG-IUS EA

DUB LNG-IUS EA

Fibroids >12 Fibroids >12 >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA

Polyp / SMF >12 by

bimanual exam


<12 by

bimanual exam


endometrium LNG-IUS

inserted

EA

322

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


Fibroids <12 >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA

ED ED Hyperplasia LNG-IUS Hysterectomy

Cancer Hysterectomy GP appt


shows hyperplasia /


hysterectomy

Hysterectomy /

GP visit

Fibroids <12 LNG-IUS Hysterectomy as

EBx prior to EA

shows ED


EBx prior to EA

shows ED

DUB DUB LNG-IUS EA


endometrium LNG-IUS

inserted

EA

Fibroids <12 LNG-IUS EA

ED Normal

endometrium

LNG-IUS EA

EBx alone Polyp / SMF Polyp / SMF Resection LNG-IUS


diagnosed at pre-

ablation

hysteroscopy

Complex

hyperplasia

LNG-IUS Hysterectomy

323

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


Fibroids <12 Normal cavity LNG-IUS EA

Polyp / SMF Resection but no lesion

seen so LNG-IUS

inserted

EA

Complex

hyperplasia


Fibroids > 12 Normal cavity >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA

Polyp / SMF >12 by

bimanual exam


<12 by

bimanual exam

Resection but no lesion

seen so LNG-IUS

inserted

EA

Complex

hyperplasia

>12 by

bimanual exam


<12 by

bimanual exam


ED Complex

hyperplasia


Atypia /

cancer



shows hyperplasia /


hysterectomy

Hysterectomy /

GP visit


BX prior to EA

shows ED

324

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


DUB Normal cavity LNG-IUS EA


endometrium LNG-IUS

inserted

EA

OPH and EBx Polyp / SMF Polyp / SMF Polyp / SMF Resection LNG-IUS

DUB Resection LNG-IUS

Complex

hyperplasia

Resection followed by

LNG-IUS

Hysterectomy

ED Polyp / SMF LNG-IUS Resection- lesion

diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy

Complex

hyperplasia


Normal cavity Polyp / SMF LNG-IUS Resection- lesion

diagnosed at pre-

ablation

hysteroscopy

DUB LNG-IUS EA

Complex

hyperplasia


Fibroids <12 DUB DUB LNG-IUS EA

Polyp / SMF LNG-IUS EA

Complex

hyperplasia


325

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


Polyp / SMF DUB Resection but as normal

endometrium LNG-IUS

inserted

EA


endometrium LNG-IUS

inserted

EA

Complex

hyperplasia


LNG-IUS

Hysterectomy

ED DUB LNG-IUS EA


Complex

hyperplasia


Fibroids >12 DUB DUB >12 by bimanual exam Hysterectomy GP appt

<12 by bimanual exam LNG-IUS EA

Polyp / SMF >12 by bimanual exam Hysterectomy GP appt


Complex

Hyperplasia

>12 by bimanual exam Hysterectomy GP appt

<12 by bimanual exam LNG-IUS Hysterectomy

Polyp / SMF DUB >12 by bimanual exam Hysterectomy GP appt

<12 by bimanual exam Resection but as normal

endometrium LNG-IUS

inserted

EA



endometrium LNG-IUS

inserted

EA

Complex

Hyperplasia


<12 by bimanual exam Resection followed by

LNG-IUS

Hysterectomy

326

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


ED DUB >12 by bimanual exam Hysterectomy GP appt




Complex

Hyperplasia



ED ED Complex

hyperplasia


Cancer / atypia Hysterectomy GP appt

Polyp /SMF LNG-IUS Hysterectomy as

pre ablation EBx

shows ED


pre ablation EBx

shows ED

Polyp / SMF Complex

hyperplasia


LNG-IUS

Hysterectomy


Polyp /SMF Resection but histology

shows complex

hyperplasia / cancer so

LNG-IUS/hysterectomy

Hysterectomy /

GP appt

DUB Resection but histology

shows complex



Hysterectomy /

GP appt

DUB Complex

hyperplasia



327

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


Polyp /SMF LNG-IUS Hysterectomy as

pre ablation EBx

shows ED


pre ablation EBx

shows ED

DUB DUB DUB LNG-IUS EA

Polyp /SMF LNG-IUS EA

Polyp/SMF DUB Resection followed by

LNG-IUS

EA

Polyp /SMF Resection but as normal

endometrium LNG-IUS

inserted

EA

ED DUB LNG-IUS EA

Polyp /SMF LNG-IUS EA

SIS and EBx

and

TVS and EBx

Polyp/SMF Polyp/SMF Polyp/SMF Resection LNG-IUS


Complex

hyperplasia


LNG-IUS

Hysterectomy

Fibroids <12 Polyp/SMF Resection LNG-IUS


diagnosed at pre-

ablation

hysteroscopy

Complex

hyperplasia


328

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


ED Polyp/SMF Resection LNG-IUS


diagnosed at pre-

ablation

hysteroscopy

Complex

hyperplasia


DUB Polyp/SMF LNG IUS Resection- lesion

diagnosed at pre-

ablation

hysteroscopy

LNG-IUS


diagnosed at pre-

ablation

Complex

hyperplasia


Fibroids <12 Fibroids <12 DUB LNG-IUS EA

Polyp/SMF Resection but as normal

endometrium LNG-IUS

inserted

EA

Complex

hyperplasia



endometrium LNG-IUS

inserted

EA


endometrium LNG-IUS

inserted

EA

329

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


Complex

hyperplasia


endometrium LNG-IUS

inserted

Hysterectomy

ED DUB LNG-IUS EA


endometrium LNG-IUS

inserted

EA

Complex

hyperplasia


DUB DUB LNG-IUS EA

Polyp/SMF LNG-IUS EA

Complex

hyperplasia


Fibroids >12 Fibroid >12 DUB >12 by bimanual exam Hysterectomy GP appt


Polyp/SMF >12 by bimanual exam Hysterectomy GP appt


endometrium LNG-IUS

inserted

EA

Complex

hyperplasia





endometrium LNG-IUS

inserted

EA



endometrium LNG-IUS

inserted

EA

330

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


Complex

hyperplasia


<12 by bimanual exam Resection followed by

LNG-IUS

Hysterectomy

Fibroid <12 DUB >12 by bimanual exam Hysterectomy GP appt




endometrium LNG-IUS

inserted

EA

Complex

hyperplasia



ED ED Complex

hyperplasia



Polyp / SMF Resection but histology

shows complex



Hysterectomy /

GP appt


pre-ablation

testing shows ED

Polyp / SMF Complex

hyperplasia


LNG-IUS

Hysterectomy

Cancer / atypia Resection followed by

hysterectomy

GP appt


shows complex



Hysterectomy /

GP appt

331

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2



shows complex



Hysterectomy /

GP appt

Fibroids < 12 Complex

hyperplasia




shows complex



Hysterectomy /

GP appt


pre-ablation

testing shows ED

DUB Complex

hyperplasia



Polyp / SMF LNG-IUS Hysterectomy as

pre-ablation

testing shows ED


pre-ablation

testing shows ED



Polyps / SMF DUB Resection but as normal

endometrium LNG-IUS

inserted

EA


endometrium LNG-IUS

inserted

EA

332

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


Fibroids < 12 DUB LNG-IUS EA


endometrium LNG-IUS

inserted

EA

ED DUB LNG-IUS EA


endometrium LNG-IUS

inserted

EA

SIS and OPH

and

TVS and OPH

Polyp / SMF Polyp / SMF Polyp / SMF Resection LNG-IUS

ED Polyp/SMF LNG-IUS Resection- lesion

diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy

Fibroid <12 Polyp / SMF Resection LNG-IUS


diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy

333

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


ED Polyp / SMF Resection LNG-IUS


diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy

DUB Polyp / SMF Resection LNH-IUS


diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy

Fibroids <12 Fibroids <12 DUB LNG-IUS EA


endometrium LNG-IUS

inserted

EA

ED Normal endometrium LNG-IUS EA

Polyp / SMF DUB LNG-IUS EA


endometrium LNG-IUS

inserted

EA


ED DUB Normal endometrium LNG-IUS EA

334

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2



endometrium LNG-IUS

inserted

EA


DUB DUB LNG-IUS EA


endometrium LNG-IUS

inserted

EA


Fibroids >12 Fibroids > 12 DUB >12 by bimanual exam Hysterectomy GP appt




endometrium LNG-IUS

inserted

EA

ED >12 by bimanual exam Hysterectomy GP appt

<12 by bimanual exam Normal

endometrium

LNG-IUS EA





endometrium LNG-IUS

inserted

EA



endometrium

LNG-IUS EA

Fibroid <12 DUB >12 by bimanual exam Hysterectomy GP appt


335

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2




endometrium LNG-IUS

inserted

EA



endometrium

LNG-IUS EA

ED ED ED Complex hyperplasia LNG-IUS Hysterectomy



shows complex



Hysterectomy /

GP appt

DUB Complex hyperplasia LNG-IUS Hysterectomy


Polyp / SMF ED Complex hyperplasia LNG-IUS Hysterectomy



shows complex



Hysterectomy /

GP appt


pre-ablation

testing shows ED

Fibroid <12 ED Complex hyperplasia LNG-IUS Hysterectomy



shows complex



Hysterectomy /

GP appt

336

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2



pre-ablation

testing shows ED

DUB ED Complex hyperplasia LNG-IUS Hysterectomy



shows complex



Hysterectomy /

GP appt


pre-ablation

testing shows ED



endometrium LNG-IUS

inserted

EA


Polyp / SMF DUB LNG-IUS EA


endometrium LNG-IUS

inserted

EA


Fibroids <12 DUB LNG-IUS


endometrium LNG-IUS

inserted

EA


ED DUB Normal endometrium LNG-IUS EA


endometrium LNG-IUS

inserted

EA


337

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


SIS and OPH

and EBx

and

TVS and OPH

and EBx

Polyp / SMF Polyp / SMF Polyp / SMF Polyp / SMF Resection LNG-IUS


Complex hyperplasia Resection followed by

LNG-IUS

Hysterectomy


diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy

Complex hyperplasia LNG-IUS Hysterectomy

DUB Polyp / SMF LNG-IUS Resection- lesion

diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy


Fibroids <12 Polyp / SMF Polyp / SMF Resection LNG-IUS



LNG-IUS

Hysterectomy

338

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2



diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy



diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy


ED Polyp / SMF Polyp / SMF Resection LNG-IUS



LNG-IUS

Hysterectomy


diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy


339

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2



diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy


DUB Polyp / SMF Polyp / SMF Resection LNG-IUS



LNG-IUS

Hysterectomy


diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy



diagnosed at pre-

ablation

hysteroscopy


diagnosed at pre-

ablation

hysteroscopy


340

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2





endometrium LNG-IUS

inserted

EA


endometrium LNG-IUS

inserted

EA


LNG-IUS

Hysterectomy

ED DUB LNG-IUS EA



Polyp / SMF DUB DUB LNG-IUS EA




endometrium LNG-IUS

inserted

EA


endometrium LNG-IUS

inserted

EA


LNG-IUS

Hysterectomy

ED DUB LNG-IUS EA



ED DUB DUB LNG-IUS EA



341

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2



endometrium LNG-IUS

inserted

EA


endometrium LNG-IUS

inserted

EA


LNG-IUS

Hysterectomy

ED DUB LNG-IUS EA







endometrium LNG-IUS

inserted

EA


endometrium LNG-IUS

inserted

EA


LNG-IUS

Hysterectomy

ED DUB LNG-IUS EA



Fibroids > 12 Fibroids > 12 DUB DUB >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA

Polyp / SMF >12 by

bimanual exam


342

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


<12 by

bimanual exam

LNG-IUS EA

Complex hyperplasia >12 by

bimanual exam


<12 by

bimanual exam


Polyp / SMF DUB >12 by

bimanual exam


<12 by

bimanual exam


endometrium LNG-IUS

inserted

EA

Polyp / SMF >12 by

bimanual exam


<12 by

bimanual exam


endometrium LNG-IUS

inserted

EA


bimanual exam


<12 by

bimanual exam


LNG-IUS

Hysterectomy

ED DUB >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA

Polyp / SMF >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA


bimanual exam


<12 by

bimanual exam


343

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


Polyp / SMF DUB DUB >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA

Polyp / SMF >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA


bimanual exam


<12 by

bimanual exam



bimanual exam


<12 by

bimanual exam

Resection but as

normal endometrium

LNG-IUS inserted

EA

Polyp / SMF >12 by

bimanual exam


<12 by

bimanual exam


endometrium LNG-IUS

inserted

EA


bimanual exam


<12 by

bimanual exam


LNG-IUS

Hysterectomy

ED DUB >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA

344

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2


Polyp / SMF >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA


bimanual exam


<12 by

bimanual exam


Fibroids < 12 DUB DUB >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA

Polyp / SMF >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA


bimanual exam


<12 by

bimanual exam



bimanual exam


<12 by

bimanual exam


endometrium LNG-IUS

inserted

EA

Polyp / SMF >12 by

bimanual exam


<12 by

bimanual exam


endometrium LNG-IUS

inserted

EA

345

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2



bimanual exam


<12 by

bimanual exam


LNG-IUS

Hysterectomy

ED DUB >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA

Polyp / SMF >12 by

bimanual exam


<12 by

bimanual exam

LNG-IUS EA


bimanual exam


<12 by

bimanual exam


ED ED ED Complex hyperplasia LNG-IUS Hysterectomy



pre-ablation

testing shows ED


pre-ablation

testing shows ED

Polyp / SMF Complex hyperplasia Removal followed by

LNG-IUS

Hysterectomy

Cancer / atypia Removal followed by

hysterectomy

GP appt

346

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2



shows complex



Hysterectomy /

GP appt


shows complex



Hysterectomy /

GP appt




pre-ablation

testing shows ED


pre-ablation

testing shows ED

Polyp / SMF ED Complex hyperplasia LNG-IUS Hysterectomy



pre-ablation

testing shows ED


pre-ablation

testing shows ED

Polyp / SMF Complex hyperplasia LNG-IUS Hysterectomy



shows complex



Hysterectomy /

GP appt

347

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2



shows complex



Hysterectomy /

GP appt




pre-ablation

testing shows ED


pre-ablation

testing shows ED

Fibroids <12 ED Complex hyperplasia LNG-IUS Hysterectomy



pre-ablation

testing shows ED


pre-ablation

testing shows ED

Polyp / SMF Complex hyperplasia Removal followed by

LNG-IUS

Hysterectomy

Cancer / atypia Removal followed by

hysterectomy

GP appt


shows complex



Hysterectomy /

GP appt

348

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2



shows complex



Hysterectomy /

GP appt




pre-ablation

testing shows ED


pre-ablation

testing shows ED

DUB ED Complex hyperplasia LNG-IUS Hysterectomy



pre-ablation

testing shows ED


pre-ablation

testing shows ED

Polyp / SMF Complex hyperplasia LNG-IUS Hysterectomy



shows complex



Hysterectomy /

GP appt


shows complex



Hysterectomy /

GP appt

349

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2





pre-ablation

testing shows ED


pre-ablation

testing shows ED

DUB DUB DUB DUB LNG-IUS EA



endometrium LNG-IUS

inserted

EA


endometrium LNG-IUS

inserted

EA

ED DUB LNG-IUS EA


Polyp / SMF DUB DUB LNG-IUS EA



endometrium LNG-IUS

inserted

EA


endometrium LNG-IUS

inserted

EA

ED DUB LNG-IUS EA


Fibroid <12 DUB DUB LNG-IUS EA


350

A6.2 continued

Strategy True

pathology

Diagnosis

from test 1

Diagnosis

from test 2



endometrium LNG-IUS

inserted

EA


endometrium LNG-IUS

inserted

EA

ED DUB LNG-IUS EA


ED DUB DUB LNG-IUS EA



endometrium LNG-IUS

inserted

EA


endometrium LNG-IUS

inserted

EA

ED DUB LNG-IUS EA


351

APPENDIX 7

Search strategies for collection of data to populate the decision trees for the

economic analysis of heavy menstrual bleeding

A7.1a Medline search strategy for prevalence of endometrial polyps

1 endometrial.ti,ab

2 endometr*.ti,ab

3 uterine.ti,ab

4 uter*.ti,ab

5 exp UTERINE DISEASES/

6 uterus.ti,ab

7 exp UTERUS/

8 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7

9 polyp.ti,ab

10 polyp*.ti,ab

11 exp POLYPS/

12 9 OR 10 OR 11

13 8 AND 12

14 hysteroscopy.ti,ab

15 exp HYSTEROSCOPY/

16 hysteroscop*.ti,ab

17 14 OR 15 OR 16

18 sensitiv*.ti,ab

19 exp "SENSITIVITY AND SPECIFICITY"/

20 diagnos*.ti,ab

21 DIAGNOSIS/

22 diagnostic*.ti,ab

23 DIAGNOSIS, DIFFERENTIAL/

24 *DIAGNOSIS/

25 18 OR 19 OR 20 OR 21 OR 22 OR 23 OR 24

26 13 AND 17 AND 25

352

A7.1b Embase search strategy for prevalence of endometrial polyps

1 endometrial.ti,ab

2 endometr*.ti,ab

3 exp ENDOMETRIAL DISEASE/

4 uter*.ti,ab

5 uterine.ti,ab

6 exp UTERUS/

7 uterus.ti,ab

8 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7

9 polyp.ti,ab

10 polyp*.ti,ab

11 exp POLYP/ OR exp ENDOMETRIUM POLYP/

12 9 OR 10 OR 11

18 8 AND 12




22 19 OR 20 OR 21

23 18 AND 22

24 sensitiv*.ti,ab

25 exp "SENSITIVITY AND SPECIFICITY"/

26 diagnos*.ti,ab

27 DIAGNOSIS/

28 DIFFERENTIAL DIAGNOSIS/

29 DIAGNOSTIC TEST/

30 24 OR 25 OR 26 OR 27 OR 28 OR 29

31 18 AND 30

32 22 AND 31

353

A7.2a Medline search strategy for prevalence of fibroids

1 prevalence.ti,ab

2 exp PREVALENCE/

3 1 OR 2

4 exp LEIOMYOMA/

5 fibroid.ti,ab

6 leiomyoma.ti,ab

7 exp MYOMA/

8 fibromyoma.ti,ab

9 leiofibromyoma.ti,ab

10 fibroleiomyoma.ti,ab

11 fibroma.ti,ab

12 exp FIBROMA/

13 myoma*.ti,ab

14 4 OR 5 OR 6 OR 7 OR 8 OR 9 OR 10 OR 11 OR 12 OR 13

15 sonogr*.ti,ab

16 hysterosonogr*.ti,ab

17 ultrasound.ti,ab

18 exp ULTRASONOGRAPHY/

19 ((transvaginal scan)).ti,ab

20 hysterosco*.ti,ab


22 sonohyster*.ti,ab

23 15 OR 16 OR 17 OR 18 OR 19 OR 20 OR 21 OR 22

24 3 AND 14 AND 23

354

A7.2b Embase search strategy for prevalence of fibroids

1 prevalence.ti,ab

2 exp PREVALENCE/

3 fibroid*.ti,ab

4 exp LEIOMYOMA/ OR exp UTERUS MYOMA/

5 leiomyoma.ti,ab

6 myoma*.ti,ab

7 exp MYOMA/

8 fibromyoma.ti,ab

9 leiofibromyoma.ti,ab

10 fibroleiomyoma.ti,ab

11 fibroma.ti,ab

12 exp FIBROMA/

13 1 OR 2


15 sonogra*.ti,ab

16 hysterosonogra*.ti,ab

17 sonohyster*.ti,ab

18 ultrasound.ti,ab

19 exp ULTRASOUND/

20 ((transvaginal scan)).ti,ab

21 hysterosco*.ti,ab


23 15 OR 16 OR 17 OR 18 OR 19 OR 20 OR 21 OR 22

24 13 AND 14 AND 23

355

A7.3a Medline search strategy for the prevalence of endometrial hyperplasia

1 exp HEMORRHAGE/

2 bleeding.ti,ab

3 exp BLOOD/

4 blood.ti,ab

5 exp MENORRHAGIA/

6 menorrhagia.ti,ab

7 menstr*.ti,ab

8 exp MENSTRUAL CYCLE/

9 1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7 OR 8

10 prevalence.ti,ab

11 exp PREVALENCE/

12 10 OR 11

24 uterine.ti,ab

25 uterus.ti,ab

26 exp UTERUS/

27 uter*.ti,ab

28 endometrial.ti,ab

29 exp ENDOMETRIAL HYPERPLASIA/

30 endometr*.ti,ab

31 hyperplas*.ti,ab

32 hyperplasia.ti,ab

33 exp HYPERPLASIA/

34 24 OR 25 OR 26 OR 27 OR 28 OR 30

35 29 OR 31 OR 32 OR 33

37 9 AND 12 AND 34 AND 35

356

A7.3b Embase search strategy for the prevalence of endometrial hyperplasia

1 hemorrhage.ti,ab

2 exp BLEEDING/

3 bleed*.ti,ab

4 blood*.ti,ab

5 exp BLOOD/

6 bleeding.ti,ab

7 exp MENORRHAGIA/

8 menorrhagia.ti,ab

9 menstr*.ti,ab



12 prevalence.ti,ab

13 exp PREVALENCE/

14 12 OR 13

15 uterus.ti,ab

16 exp UTERUS/

17 uter*.ti,ab

18 uterine.ti,ab



21 endometr*.ti,ab

22 hyperplas*.ti,ab

23 exp HYPERPLASIA/ OR exp ENDOMETRIUM HYPERPLASIA/

24 15 OR 16 OR 17 OR 18 OR 19 OR 21

25 20 OR 22 OR 23

26 24 AND 25

27 11 AND 15 AND 27

357

A7.4a Medline search strategy for the prevalence of endometrial cancer

1 exp HEMORRHAGE/

2 bleeding.ti,ab

3 exp BLOOD/

4 blood.ti,ab

5 exp MENORRHAGIA/

6 menorrhagia.ti,ab

7 menstr*.ti,ab


9 6 OR 7 OR 8 OR 9 OR 10 OR 11 OR 12 OR 13

10 prevalence.ti,ab

11 exp PREVALENCE/

12 15 OR 16

13 uterine.ti,ab

14 uterus.ti,ab

15 exp UTERUS/

16 uter*.ti,ab


19 endometr*.ti,ab

20 cancer.ti,ab

21 exp NEOPLASMS/

22 malignan*.ti,ab

23 ((Endometrial cancer)).ti,ab

24 exp ENDOMETRIAL NEOPLASMS/

25 13 OR 14 OR 15 OR 16 OR 17 OR 19

26 20 OR 21 OR 22 OR 23 OR 24

27 9 AND 12 AND 25 AND 26

358

A7.4b Embase search strategy for the prevalence of endometrial cancer

1 hemorrhage.ti,ab

2 exp BLEEDING/

3 bleed*.ti,ab

4 blood*.ti,ab

5 exp BLOOD/

6 bleeding.ti,ab

7 exp MENORRHAGIA/

8 menorrhagia.ti,ab

9 menstr*.ti,ab



12 prevalence.ti,ab

13 exp PREVALENCE/

14 12 OR 13

15 uterus.ti,ab

16 exp UTERUS/

17 uter*.ti,ab

18 uterine.ti,ab

19 cancer.ti,ab

20 exp NEOPLASM/

21 malignan*.ti,ab

22 (endometrial AND cancer).ti,ab



25 endometr*.ti,ab

26 15 OR 16 OR 17 OR 18 OR 23 OR 24 OR 25

27 19 OR 20 OR 21

28 26 AND 27

29 22 OR 28

30 11 AND 14 AND 29

359

A7.5a Medline search strategy for reviews of outpatient hysteroscopy

1 exp HYSTEROSCOPY/



4 1 OR 2 OR 3

5 4 [Limit to: Review Articles]

A7.5b Embase search strategy for reviews of outpatient hysteroscopy

1 exp HYSTEROSCOPY/



4 1 OR 2 OR 3

5 4 [Limit to: (Publication Types Review)]

360

A7.6a Medline search strategy for studies of transvaginal ultrasound and heavy

menstrual bleeding

1 (transvaginal AND ultrasound).ti,ab

2 exp ULTRASONOGRAPHY/

3 sonogra*.ti,ab

4 transvaginal.ti,ab

5 vaginal.ti,ab

6 2 OR 3

7 4 OR 5

8 6 AND 7

9 1 OR 8

10 "abnormal uterine bleeding".ti,ab

11 exp METRORRHAGIA/ OR exp MENORRHAGIA/

12 menorrhagia.ti,ab

13 10 OR 11 OR 12

14 9 AND 13

A7.6b Embase search strategy for studies of transvaginal ultrasound and heavy

menstrual bleeding

1 (transvaginal AND ultrasound).ti,ab

2 exp ULTRASOUND/

3 sonogra*.ti,ab

4 transvaginal.ti,ab

5 vaginal.ti,ab

6 4 OR 5

7 2 OR 3

8 6 AND 7

9 1 OR 8

10 exp MENORRHAGIA/ OR exp UTERUS BLEEDING/


12 menorrhagia.ti,ab

13 10 OR 11 OR 12

14 9 AND 13

361

A7.7a Medline search strategy for studies of saline infusion sonography and heavy

menstrual bleeding

1 exp MENORRHAGIA/

2 menorrhagia.ti,ab

3 hypermenorrhea.ti,ab

4 (heavy ADJ menstrual ADJ bleeding).ti,ab

5 (heavy ADJ menstrua*).ti,ab


7 1 OR 2 OR 3 OR 4 OR 5 OR 6

8 exp SODIUM CHLORIDE/ AND exp ULTRASONOGRAPHY/

9 (saline AND infusion AND sonography).ti,ab

10 (sono AND hysterosonography).ti,ab

11 (saline AND hysterosonography).ti,ab

12 (saline AND hysterography).ti,ab

13 sonohysterography.ti,ab

14 8 OR 9 OR 10 OR 11 OR 12 OR 13

15 7 AND 14

A7.7b Embase search strategy for studies of saline infusion sonography and heavy

menstrual bleeding

1 exp MENORRHAGIA/

2 menorrhagia.ti,ab

3 hypermenorrhea.ti,ab

4 (heavy ADJ menstrual ADJ bleeding).ti,ab

5 (heavy ADJ menstrua*).ti,ab


7 1 OR 2 OR 3 OR 4 OR 5 OR 6

8 exp SODIUM CHLORIDE/ AND exp ULTRASONOGRAPHY/

9 (saline AND infusion AND sonography).ti,ab

10 (sono AND hysterosonography).ti,ab

11 (saline AND hysterosonography).ti,ab

12 (saline AND hysterography).ti,ab

13 sonohysterography.ti,ab

14 8 OR 9 OR 10 OR 11 OR 12 OR 13

15 7 AND 14

362

A7.8a Medline search strategy for reviews of endometrial biopsy

1 (endometrial AND biopsy).ti,ab

2 exp ENDOMETRIUM/

3 endometr*.ti,ab

4 exp BIOPSY/

5 biopsy.ti,ab

6 sampling.ti,ab

7 2 OR 3

8 4 OR 5 OR 6

9 7 AND 8

10 1 OR 9

11 10 [Limit to: Review Articles]

A7.8b Embase search strategy for reviews of endometrial biopsy

1 (endometrial AND biopsy).ti,ab

2 exp ENDOMETRIUM BIOPSY/

3 exp ENDOMETRIUM/

4 endometr*.ti,ab

5 biopsy.ti,ab

6 exp BIOPSY/

7 sampling.ti,ab

8 exp SAMPLING/

9 5 OR 6 OR 7 OR 8

10 3 OR 4

11 9 AND 10

12 1 OR 2 OR 11

13 12 [Limit to: (Publication Types Review)]

363

A7.9a Medline search strategy for studies of LNG-IUS for heavy menstrual bleeding

1 menorrhag*.ti,ab

2 exp MENORRHAGIA/

3 "heavy menstrual blee*".ti,ab

4 menometrorrhagia.ti,ab

5 METRORRHAGIA/

6 hypermenorrh*.ti,ab

7 1 OR 2 OR 3 OR 4 OR 5 OR 6

8 mirena.ti,ab

9 exp LEVONORGESTREL/

10 "intrauterine device".ti,ab

11 "intrauterine system".ti,ab

12 INTRAUTERINE DEVICES, MEDICATED/ OR INTRAUTERINE

DEVICES/ 13 IUS.ti,ab

14 IUD.ti,ab

15 LNG-IUS.ti,ab

16 levonorgestrel-releasing.ti,ab

17 8 OR 9 OR 10 OR 11 OR 12 OR 13 OR 14 OR 15 OR 16

18 7 AND 17

A7.9b Embase search strategy for studies of LNG-IUS for heavy menstrual bleeding

1 menorrhag*.ti,ab

2 exp MENORRHAGIA/

3 "heavy menstrual bleed*".ti,ab

4 exp MENSTRUATION DISORDER/

5 menometorrhagia.ti,ab

6 hypermenorrh*.ti,ab

7 1 OR 2 OR 3 OR 4 OR 5 OR 6

8 mirena.ti,ab

9 exp LEVONORGESTREL/

10 "intrauterine system".ti,ab

11 IUS.ti,ab

12 LNG-IUS.ti,ab

13 IUD.ti,ab

14 "intrauterine device".ti,ab

15 levonorgestrel-releasing.ti,ab

16 8 OR 9 OR 10 OR 11 OR 12 OR 13 OR 14 OR 15

17 7 AND 16

364

A7.10a Medline search strategy for patient satisfaction after endometrial ablation

1 "endometrial ablation".ti,ab

2 exp ENDOMETRIAL ABLATION TECHNIQUES/

3 satisfaction.ti,ab

4 exp PATIENT SATISFACTION/

5 1 OR 2

6 3 OR 4

7 5 AND 6

A7.10b Embase search strategy for patient satisfaction after endometrial ablation

1 "endometrial ablation".ti,ab

2 exp ENDOMETRIUM ABLATION/


4 exp PATIENT SATISFACTION/ OR exp SATISFACTION/

5 3 OR 4

6 1 OR 2

7 5 AND 6

A7.10c Medline search for endometrial ablation and fibroids

1 fibroid*.ti,ab

2 UTERUS MYOMA/ OR LEIOMYOMA/

3 myoma.ti,ab

4 leiomyoma.ti,ab

5 1 OR 2 OR 3 OR 4

6 (endometrial AND ablation).ti,ab

7 exp ENDOMETRIAL ABLATION TECHNIQUES/

7 6 OR 7

9 5 AND 8

365

A7.10d Embase search for endometrial ablation and fibroids

1 ((endometrial ablation)).ti,ab

2 exp ENDOMETRIUM ABLATION/

3 1 OR 2

4 fibroid*.ti,ab

5 UTERUS MYOMA/ OR LEIOMYOMA/

6 myoma.ti,ab

7 leiomyoma.ti,ab

8 4 OR 5 OR 6 OR 7

9 3 AND 8

A7.11a Medline search strategy for satisfaction with hysterectomy as a treatment for

fibroids

1 exp HYSTERECTOMY/

2 exp LEIOMYOMA/

3 exp PERSONAL SATISFACTION/


5 3 OR 4

6 1 AND 2 AND 5

A7.11b Embase search strategy for satisfaction with hysterectomy as a treatment for

fibroids

1 exp HYSTERECTOMY/

2 exp LEIOMYOMA/


4 exp SATISFACTION/

5 3 OR 4

6 1 AND 2 AND 5

366

A7.12a Medline search strategy for satisfaction after endometrial polypectomy

1 polypectomy.ti,ab

2 (endometrial AND polyp).ti,ab

3 exp POLYPS/ AND exp ENDOMETRIUM/

4 removal.ti,ab

5 1 OR 4

6 2 OR 3

7 5 AND 6

A7.12b Embase search strategy for satisfaction after endometrial polypectomy

A7.12c Medline search strategy for satisfaction after transcervical resection of a fibroid

1 (transcervical AND resection AND fibroid).ti,ab

2 exp LEIOMYOMA/

3 (hysteroscopic AND removal).ti,ab

4 myomectomy.ti,ab

5 exp GYNECOLOGIC SURGICAL PROCEDURES/

6 fibroid.ti,ab

7 (submucosal AND fibroid).ti,ab

8 submuc*.ti,ab

9 6 AND 8

10 2 AND 8

11 1 OR 3 OR 4 OR 5

12 7 OR 9 OR 10

13 11 AND 12

14 satisf*.ti,ab


16 14 OR 15

17 13 AND 16

1 polypectomy.ti,ab

2 exp POLYPECTOMY/

3 (endometrial AND polyp).ti,ab

4 exp ENDOMETRIUM POLYP/

5 removal.ti,ab

6 1 OR 2 OR 5

7 3 OR 4

8 6 AND 7

367

A7.12d Embase search strategy for satisfaction after transcervical resection of a fibroid

1 (transcervical AND resection AND fibroid).ti,ab

2 fibroid.ti,ab

3 exp UTERUS MYOMA/ OR exp LEIOMYOMA/

4 (hysteroscopic AND removal).ti,ab

5 myomectomy.ti,ab

6 exp MYOMECTOMY/

7 submuc*.ti,ab

8 (submucosal AND fibroid).ti,ab

9 2 OR 3

10 7 AND 9

11 8 OR 10

12 1 OR 4 OR 5 OR 6

13 11 AND 12

14 satisf*.ti,ab


16 exp SATISFACTION/ OR exp PATIENT SATISFACTION/

17 14 OR 15 OR 16

18 13 AND 17

A7.13a Medline search strategy for satisfaction after dilatation and curettage

1 D+C.ti,ab

2 exp "DILATATION AND CURETTAGE"/

3 curettage.ti,ab

4 CURETTAGE/

5 ((heavy menstrual bleeding)).ti,ab

6 exp MENORRHAGIA/

7 5 OR 6

8 1 OR 2 OR 3 OR 4

9 7 AND 8

368

A7.13b Embase search strategy for satisfaction after dilatation and curettage

1 D+C.ti,ab

2 exp "DILATATION AND CURETTAGE"/

3 curettage.ti,ab

4 CURETTAGE/

5 ((heavy menstrual bleeding)).ti,ab

6 exp MENORRHAGIA/

7 5 OR 6

8 1 OR 2 OR 3 OR 4

9 7 AND 8

369

A7.14a Medline search strategy for satisfaction after uterine artery embolisation

1 "uterine artery embolis*".ti,ab

2 "uterine artery emboliz*".ti,ab

3 UAE.ti,ab

4 exp UTERINE ARTERY EMBOLIZATION/

5 1 OR 2 OR 3 OR 4


7 satisf*.ti,ab

8 6 OR 7

9 5 AND 8

A7.14b Embase search strategy for satisfaction after uterine artery embolisation

1 "uterine artery embolis*".ti,ab

2 "uterine artery emboliz*".ti,ab

3 UAE.ti,ab

4 exp UTERINE ARTERY EMBOLIZATION/

5 1 OR 2 OR 3 OR 4


7 exp SATISFACTION/ OR exp PATIENT SATISFACTION/

8 satisf*.ti,ab

9 6 OR 7 OR 8

10 5 AND 9

370

A7.15a Medline search strategy for satisfaction after myomectomy

1 myomectomy.ti,ab

2 satisf*.ti,ab

3 2 AND 3

A7.15b Embase search strategy for satisfaction after myomectomy

1 myomectomy.ti,ab

2 satisf*.ti,ab

3 2 AND 3

371

REFERENCES

(1) Royal College of Obstetricians and Gynaecologists. Standards for Gynaecology.

Report of a working party. London: RCOG Press. 2008.

(2) Gulumser C, Narvekar N, Pathak M, Palmer E, Parker S, Saridogan E. See-and-treat

outpatient hysteroscopy: An analysis of 1109 examinations. Reproductive

BioMedicine Online 2010;20(3):423-9.

(3) Maiti S, Naidoo K. Patients' satisfaction survey at the outpatient hysteroscopy

service at St. Mary's hospital, Manchester, UK. Menopause International

2008;14(4).

(4) Nagele F, O'Connor H, Davies A, Badawy A, Mohamed H, Magos A. 2500

Outpatient diagnostic hysteroscopies. Obstetrics & Gynecology 1996;88(1):87-92.

(5) Clark TJ, Mahajan D, Sunder P, Gupta JK. Hysteroscopic treatment of symptomatic

submucous fibroids using a bipolar intrauterine system: a feasibility study. European

Journal of Obstetrics, Gynecology, & Reproductive Biology 2002;100(2):237-42.

(6) Clark TJ, Gupta JK. Handbook of Outpatient Hysteroscopy. A complete guide to

diagnosis and therapy. First ed. London: Hodder Education; 2005.

(7) Moore K.L, Dalley A.F. Pelvis and Perineum. In: Paul J Kelly, editor. Clinically

oriented anatomy. Fourth ed. Baltimore: Lippincott Williams and Wilkins; 1992. p.

374-5.

(8) Finikiotis G. Side-effects and complications of outpatient hysteroscopy. Australian

& New Zealand Journal of Obstetrics & Gynaecology 1993;33:61-2.

(9) Cicinelli E, Didonna T, Ambrosi G, Schonauer LM, Fiore G, Matteo MG. Topical

anaesthesia for diagnostic hysteroscopy and endometrial biopsy in postmenopausal

women: a randomised placebo-controlled double-blind study. British Journal of

Obstetrics & Gynaecology 1997;104(3):316-9.

(10) Cicinelli E, Didonna T, Schonauer LM, Stragapede S, Falco N, Pansini N.

Paracervical anesthesia for hysteroscopy and endometrial biopsy in postmenopausal

women. A randomized, double-blind, placebo-controlled study. Journal of

Reproductive Medicine 1998;43(12):1014-8.

(11) Zupi E, Luciano AA, Valli E, Marconi D, Maneschi F, Romanini C. The use of

topical anesthesia in diagnostic hysteroscopy and endometrial biopsy. Fertility and

Sterility 1995;63(2):414-6.

(12) Lau WC, Lo WK, Tam WH, Yuen PM. Paracervical anaesthesia in outpatient

hysteroscopy: a randomised double-blind placebo-controlled trial. British Journal of

Obstetrics & Gynaecology 1999;106(4):356-9.

372

(13) Yang J, Vollenhoven B. Pain control in outpatient hysteroscopy. Obstetrical and

Gynecological Survey 2002;57(10):693-702.

(14) Tangsiriwatthana T, Sangkomkamhang US, Lumbiganon P, Laopaiboon M.

Paracervical local anaesthesia for cervical dilatation and uterine intervention.

Cochrane Database of Systematic Reviews 2009;(1).

(15) Intercollegiate working party chaired by the Royal College of Anaesthetists.

Implementing and ensuring safe sedation practice for healthcare procedures in

adults. UK Academy of Medical Royal Colleges and their Faculties; 2001.

(16) Jansen FW, Vredevoogd CB, van UK, Hermans J, Trimbos JB, Trimbos-Kemper

TC. Complications of hysteroscopy: a prospective, multicenter study. Obstetrics &

Gynecology 2000;96(2):266-70.

(17) Pasini A, Belloni C. Intraoperative complications of 697 consecutive operative

hysteroscopies [Italian] Complicanze intraoperatorie di 697 isteroscopie operative

consecutive. Minerva Ginecologica 2001;53(1):13-20.

(18) Darwish AM, Ahmad AM, Mohammad AM. Cervical priming prior to operative

hysteroscopy: A randomized comparison of laminaria versus misoprostol. Human

Reproduction 2004;19(10):2391-4.

(19) Fernandez H, Alby JD, Tournoux C, Chauveaud-Lambling A, de TR, Frydman R, et

al. Vaginal misoprostol for cervical ripening before operative hysteroscopy in pre-

menopausal women: A double-blind, placebo-controlled trial with three dose

regimens. Human Reproduction 2004;19(7):1618-21.

(20) Fung TM, Lam MH, Wong SF, Ho LC. A randomised placebo-controlled trial of

vaginal misoprostol for cervical priming before hysteroscopy in postmenopausal

women. BJOG: An International Journal of Obstetrics and Gynaecology

2002;109(5):561-5.

(21) Healey S, Butler B, Kum FN, Dunne J, Hutchens D, Crane JM. A randomized trial

of oral misoprostol in premenopausal women before hysteroscopy. Journal of

Obstetrics & Gynaecology Canada: JOGC 2007;29(8):648-52.

(22) Ngai SW, Chan YM, Liu KL, Ho PC. Oral misoprostol for cervical priming in non-

pregnant women. Human Reproduction 1997;12(11):2373-5.

(23) Broekhuizen FF. Misoprostol, its use in obstetrics and gynecology. Current

Problems in Obstetrics, Gynecology and Fertility 2000;23(6):196-220.

(24) Crane JM, Healey S. Use of misoprostol before hysteroscopy: a systematic review.

Journal of Obstetrics & Gynaecology Canada: JOGC 2006;28(5):373-9.

(25) British National Formulary. 57 ed. London: BMJ group and RPS publishing; 2009.

(26) Fernandez H, Durlot F, Dubois C, Ulmann A, Frydman R. [The role of the

progesterone antagonist, RU 486 (mifepristone), in preoperative cervical dilatation

373

during the luteal phase. A randomized double-blind study]. [French] Place d'un

antagoniste de la progesterone, RU 486 (mifepristone), en phase luteale avant la

chirurgie endo-uterine. Resultat d'un essai therapeutique en double aveugle.

Contraception, Fertilite, Sexualite 1990;18(1):41-3.

(27) Coloma F, Diago VJ, Lopez-Olmos J, Costa S, Paya V, Valero V, et al. Use of

Dilapan stem in surgical hysteroscopy. Progresos en Obstetricia y Ginecologia

1997;40(8):554-8.

(28) Townsend DE, Melkonian R. Laminaria tent for diagnostic and operative

hysteroscopy. Journal of Gynecologic Surgery 1990;6(4):271-4.

(29) Busquets M, Lemus M. [Practicability of panoramic hysteroscopy with CO2.

Clinical experience: 923 cases]. [Spanish] Factibilidad de histeroscopia panoramica

con CO2. Experiencia clinica: 923 casos. Revista Chilena de Obstetricia y

Ginecologia 1993;58(2):113-8.

(30) Bettocchi S, Selvaggi L. A vaginoscopic approach to reduce the pain of office

hysteroscopy. Journal of the American Association of Gynecologic Laparoscopists

1997;4(2):255-8.

(31) Garbin O, Kutnahorsky R, Gollner JL, Vayssiere C. Vaginoscopic versus

conventional approaches to outpatient diagnostic hysteroscopy: A two-centre

randomized prospective study. Human Reproduction 2006;21(11):2996-3000.

(32) Sharma M, Taylor A, Di Spiezio SA, Buck L, Mastrogamvrakis G, Kosmas I, et al.

Outpatient hysteroscopy: Traditional versus the 'no-touch' technique. BJOG: An

International Journal of Obstetrics and Gynaecology 2005;112(7):963-7.

(33) Sagiv R, Sadan O, Boaz M, Dishi M, Schechter E, Golan A. A new approach to

office hysteroscopy compared with traditional hysteroscopy: a randomized

controlled trial. Obstetrics & Gynecology 2006;108(2):387-92.

(34) Amin HK, Neuwirth RS. Operative hysteroscopy utilizing dextran as distending

medium. Clinical Obstetrics & Gynecology 1983;26(2):277-84.

(35) Paschopoulos M, Kaponis A, Makrydimas G, Zikopoulos K, Alamanos Y,

O'Donovan P, et al. Selecting distending medium for out-patient hysteroscopy. Does

it really matter? Human Reproduction 2004;19(11):2619-25.

(36) Soderstrom RM. Distending the uterus: What medium is best? Clinical Obstetrics

and Gynecology 1992;35(2):225-8.

(37) Taylor PJ, Lewinthal D, Leader A, Pattinson HA. A comparison of Dextran 70 with

carbon dioxide as the distention medium for hysteroscopy in patients with infertility

or requesting reversal of a prior tubal sterilization. Fertility and Sterility

1987;47(5):861-3.

(38) Bartsich E, Dillon TF. Carbon dioxide hysteroscopy. American Journal of Obstetrics

& Gynecology 1976;124(7):746-50.

374

(39) Pellicer A, Diamond MP. Distending media for hysteroscopy. Obstetrics and

Gynecology Clinics of North America 1988;15(1):23-8.

(40) Marsh F, Duffy S. The technique and overview of flexible hysteroscopy. Obstetrics

and Gynecology Clinics of North America 2004;31(3):655-68.

(41) Abbott J, Hawe J, Phillips G. A prospective evaluation of pain and acceptability

during outpatient flexible hysteroscopy and endometrial biopsy. Gynaecological

Endoscopy 2002;11(1):37-41.

(42) Kremer C, Barik S, Duffy S. Flexible outpatient hysteroscopy without anaesthesia:

A safe, successful and well tolerated procedure. British Journal of Obstetrics and

Gynaecology 1998;105(6):672-6.

(43) The Cochrane Collaboration. Cochrane Handbook for systematic reviews of

intervention. www cochrane-handbook org version 5.1.0 Last accessed 25/05/2013

(44) Centre for Reviews and Dissemination. Systematic Reviews: CRD's guidance for

undertaking reviews in health care. York: CRD, University of York; 2009.

(45) Floris S, Piras B, Orru M, Silvetti E, Tusconi A, Melis F, et al. Efficacy of

intravenous tramadol treatment for reducing pain during office diagnostic

hysteroscopy. Fertility and Sterility 2007;87(1):147-51.

(46) Nagele F, Lockwood G, Magos AL. Randomised placebo controlled trial of

mefenamic acid for premedication at outpatient hysteroscopy: a pilot study. British

Journal of Obstetrics & Gynaecology 1997;104(7):842-4.

(47) Hong J-Y, Kim J. Use of paracervical analgesia for outpatient hysteroscopic surgery:

A randomized, double-blind, placebo-controlled study. Ambulatory Surgery

2006;12(4):181-5.

(48) Landis JR, Koch GG. The measurement of observer agreement for categorical data.

Biometrics 1977;33(1):159-74.

(49) Moher D, Jadad AR, Nichol G, Penman M, Tugwell P, Walsh S. Assessing the

quality of randomized controlled trials: an annotated bibliography of scales and

checklists. Controlled Clinical Trials 1995;16(1):62-73.

(50) Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJM, Gavaghan DJ, et al.

Assessing the quality of reports of randomized clinical trials: Is blinding necessary?

Controlled Clinical Trials 1996;17(1):1-12.

(51) Khan KS, Daya S, Jadad A. The importance of quality of primary studies in

producing unbiased systematic reviews. Archives of Internal Medicine

1996;156(6):661-6.

(52) Al-Sunaidi M, Tulandi T. A randomized trial comparing local intracervical and

combined local and paracervical anesthesia in outpatient hysteroscopy. Journal of

Minimally Invasive Gynecology 2007;14(2):153-5.

375

(53) Bellati U, Bonaventura A, Costanza L, Zulli S, Gentile C. Tramadol hydrochloride

versus mepivacaine hydrochloride: Comparison between two analgesic procedures

in hysteroscopy. Giornale Italiano di Ostetricia e Ginecologia 1998;20(10):469-72.

(54) Broadbent JAM, Hill NCW, Molnar BG, Rolfe KJ, Magos AL. Randomized placebo

controlled trial to assess the role of intracervical lignocaine in outpatient

hysteroscopy. British Journal of Obstetrics and Gynaecology 1992;99(9):777-9.

(55) Costello MF, Horrowitz SD, Williamson M. A prospective randomized double-blind

placebo-controlled study of local anaesthetic injected through the hysteroscope for

outpatient hysteroscopy and endometrial biopsy. Gynaecological Endoscopy

1998;7(3):121-6.

(56) Davies A, Richardson RE, O'Connor H, Baskett TF, Nagele F, Magos AL.

Lignocaine aerosol spray in outpatient hysteroscopy: a randomized double-blind

placebo-controlled trial. Fertility & Sterility 1997;67(6):1019-23.

(57) Esteve M, Schindler S, Machado SB, Borges SA, Santos CR, Coutinho E. The

efficacy of intracervical lidocaine in outpatient hysteroscopy. Gynaecological

Endoscopy 2002;11(1):33-6.

(58) Finikiotis G, Tsocanos S. Outpatient hysteroscopy: a comparison of 2 methods of

local analgesia. Australian & New Zealand Journal of Obstetrics & Gynaecology

1992;32(4):373-4.

(59) Giorda G, Scarabelli C, Franceschi S, Campagnutta E. Feasibility and pain control in

outpatient hysteroscopy in postmenopausal women: a randomized trial. Acta

Obstetricia et Gynecologica Scandinavica 2000;79(7):593-7.

(60) Guida M, Pellicano M, Zullo F, Acunzo G, Lavitola G, Palomba S, et al. Outpatient

operative hysteroscopy with bipolar electrode: a prospective multicentre randomized

study between local anaesthesia and conscious sedation. Human Reproduction

2003;18(4):840-3.

(61) Kabli N, Tulandi T. A Randomized Trial of Outpatient Hysteroscopy with and

without Intrauterine Anesthesia. Journal of Minimally Invasive Gynecology

2008;15(3):308-10.

(62) Lau WC, Tam WH, Lo WK, Yuen PM. A randomised double-blind placebo-

controlled trial of transcervical intrauterine local anaesthesia in outpatient

hysteroscopy. British Journal of Obstetrics and Gynaecology 2000;107(5):610-3.

(63) Makris N, Xygakis A, Dachlythras M, Prevedourakis C, Michalas S. Mepivacaine

local cervical anesthesia for diagnostic hysteroscopy: A randomized placebo-

controlled study. Journal of Gynecologic Surgery 2001;17(1):7-11.

(64) Shankar M, Davidson A, Taub N, Habiba M. Randomised comparison of distension

media for outpatient hysteroscopy. BJOG: An International Journal of Obstetrics &

Gynaecology 2004;111(1):57-62.

376

(65) Soriano D, Ajaj S, Chuong T, Deval B, Fauconnier A, Darai E. Lidocaine spray and

outpatient hysteroscopy: Randomized placebo-controlled trial. Obstetrics and

Gynecology 2000;96(5):661-4.

(66) Vercellini P, Colombo A, Mauro F, Oldani S, Bramante T, Crosignani PG.

Paracervical anesthesia for outpatient hysteroscopy. Fertility and Sterility

1994;62(5):1083-5.

(67) Wong AY, Wong K, Tang LC. Stepwise pain score analysis of the effect of local

lignocaine on outpatient hysteroscopy: a randomized, double-blind, placebo-

controlled trial. Fertility & Sterility 2000;73(6):1234-7.

(68) Caligiani L, Pera L, Scuderi A, Ferrarello S. Analgesia for out-patients' hysteroscopy

in post-menopausal bleeding. Acta Anaesthesiologica Italica 1994;45(2

SUPPL.):251-6.

(69) Lin YH, Hwang JL, Huang LW, Chen HJ. Use of sublingual buprenorphine for pain

relief in office hysteroscopy. Journal of Minimally Invasive Gynecology

2005;12(4):347-50.

(70) Tam WH, Yuen PM. Use of diclofenac as an analgesic in outpatient hysteroscopy: A

randomized, double-blind, placebo-controlled study. Fertility and Sterility

2001;76(5):1070-2.

(71) Atay V, Duru NK, Pabuccu R, Ergun A, Tokac G, Aydin BA. Vaginal misoprostol

for cervical dilatation before operative office hysteroscopy. Gynaecological

Endoscopy 1997;6(1):47-9.

(72) Ben-Chetrit A, Eldar-Geva T, Lindenberg T, Farhat M, Shimonovitz S, Zacut D, et

al. Mifepristone does not induce cervical softening in non-pregnant women. Human

Reproduction 2004;19(10):2372-6.

(73) da Costa AR, Pinto-Neto AM, Amorim M, Costa Paiva LHS, Scavuzzi A, Schettini

J. Use of Misoprostol Prior to Hysteroscopy in Postmenopausal Women: A

Randomized, Placebo-Controlled Clinical Trial. Journal of Minimally Invasive

Gynecology 2008;15(1):67-73.

(74) Singh N, Ghosh B, Naha M, Mittal S. Vaginal misoprostol for cervical priming prior

to diagnostic hysteroscopy--efficacy, safety and patient satisfaction: a randomized

controlled trial. Archives of Gynecology and Obstetrics 2009.

(75) Valente EP, de Amorim MMR, Costa AAR, de Miranda DV. Vaginal Misoprostol

Prior to Diagnostic Hysteroscopy in Patients of Reproductive Age: A Randomized

Clinical Trial. Journal of Minimally Invasive Gynecology 2008;15(4):452-8.

(76) Waddell G, Desindes S, Takser L, Beauchemin M-C, Bessette P. Cervical Ripening

Using Vaginal Misoprostol before Hysteroscopy: A Double-blind Randomized Trial.

Journal of Minimally Invasive Gynecology 2008;15(6):739-44.

377

(77) Almeida ZM, Pontes R, Costa HL. [Evaluation of pain in diagnostic hysteroscopy by

vaginoscopy using normal saline at body temperature as distension medium: a

randomized controlled trial]. [Portuguese] Avaliacao da dor na histeroscopia

diagnostica por vaginoscopia utilizando-se, como meio de distensao, solucao salina

a temperatura corporal: Ensaio clinico randomizado. Revista Brasileira de

Ginecologia e Obstetricia 2008;30(1):25-30.

(78) Guida M, Di Spiezio SA, Acunzo G, Sparice S, Bramante S, Piccoli R, et al.

Vaginoscopic versus traditional office hysteroscopy: a randomized controlled study.

Human Reproduction 2006;21(12):3253-7.

(79) Paschopoulos M, Anastassopoulus P, Kaponis A, Avgoustatos F, Papadononpoulos

L, Lolis D. Vaginoscopic versus conventional approach to outpatient hysteroscopy

A comparative randomised study. Gynaecological Endoscopy 2000;SUPPL:51.

(80) Brusco GF, Arena S, Angelini A. Use of carbon dioxide versus normal saline for

diagnostic hysteroscopy. Fertility and Sterility 2003;79(4):993-7.

(81) Lavitola G, Guida M, Pellicano M, Acunzo G, Cirillo D, Nappi C. Options for

uterine distension during hysteroscopy. Minerva Ginecologica 2002;54(6):461-5.

(82) Litta P, Bonora M, Pozzan C, Merlin F, Sacco G, Fracas M, et al. Carbon dioxide

versus normal saline in outpatient hysteroscopy. Human Reproduction

2003;18(11):2446-9.

(83) Nagele F, Bournas N, O'Connor H, Broadbent M, Richardson R, Magos A.

Comparison of carbon dioxide and normal saline for uterine distension in outpatient

hysteroscopy. Fertility and Sterility 1996;65(2):305-9.

(84) Baxter AJ, Beck B, Phillips K. A randomized prospective trial of rigid and flexible

hysteroscopy in an outpatient setting. Gynaecological Endoscopy 2002;11(6):357-

64.

(85) Unfried G, Wieser F, Albrecht A, Kaider A, Nagele F. Flexible versus rigid

endoscopes for outpatient hysteroscopy: A prospective randomized clinical trial.

Human Reproduction 2001;16(1):168-71.

(86) Bradburn MJ, Deeks JJ, Berlin JA, Russell LA. Much ado about nothing: a

comparison of the performance of meta-analytical methods with rare events.

Statistics in Medicine 2007;26(1):53-77.

(87) RevMan 5 [computer program]. Version 5 2008.

(88) Sterne JA, Juni P, Schulz KF, Altman DG, Bartlett C, Egger M. Statistical methods

for assessing the influence of study characteristics on treatment effects in 'meta-

epidemiological' research. Statistics in Medicine 2002;21(11):1513-24.

(89) Stata Statistical software [computer program]. Version 8.2. College Station, Texas:

StataCorp LP; 2004.

378

(90) Standard Deviation Calculator. www easycalculation com/statistics/standard-

deviation php 2009 Last accessed 29/01/2009. Online source.

(91) Altman DG. Practical statistics for medical research. First ed. London: Chapman and

Hall/CRC; 1991.

(92) Berger V. Random Samples. The probability problem solver: a complete solution

guide to any textbook.Piscataway: Research and Education Association; 1996. p.

603-69.

(93) Di Spiezio Sardo A, Taylor A, Tsirkas P, Mastrogamvrakis G, Sharma M, Magos A.

Hysteroscopy: a technique for all? Analysis of 5,000 outpatient hysteroscopies.

Fertility & Sterility 2008;89(2):438-43.

(94) Goldfarb HA. Comparison of carbon dioxide with continuous-flow technique for

office hysteroscopy. Journal of the American Association of Gynecologic

Laparoscopists 1996;3(4):571-4.

(95) Perez-Medina T, Bajo JM, Martinez-Cortes L, Castellanos P, Perez I. Six thousand

office diagnostic-operative hysteroscopies. International Journal of Gynecology and

Obstetrics 2000;71(1):33-8.

(96) Quinones RG. Hysteroscopy: Choosing distending media. International Journal of

Fertility 1984;29(2):129-32.

(97) Seki M, Mettler L, Semm K. Comparison of fluid hysteroscopy and CO2

hysteroscopy. Endoscopy 1982;14(4):141-5.

(98) Siegler AM, Kemmann E. Hysteroscopy. Obstetrical & Gynecological Survey

1975;30(9):567-88.

(99) Pellicano M, Guida M, Zullo F, Lavitola G, Cirillo D, Nappi C. Carbon dioxide

versus normal saline as a uterine distension medium for diagnostic vaginoscopic

hysteroscopy in infertile patients: A prospective, randomized, multicenter study.

Fertility and Sterility 2003;79(2):418-21.

(100) Cooper NA, Smith P, Khan KS, Clark TJ. Vaginoscopic approach to outpatient

hysteroscopy: a systematic review of the effect on pain. BJOG: An International

Journal of Obstetrics & Gynaecology 2010;117(5):532-9.

(101) Cooper NA, Smith P, Khan KS, Clark TJ. Does cervical preparation before

outpatient hysteroscopy reduce women's pain experience? A systematic review.

BJOG: An International Journal of Obstetrics & Gynaecology 2011;118(11):1292-

301.

(102) Cooper NA, Smith P, Khan KS, Clark TJ. A systematic review of the effect of the

distension medium on pain during outpatient hysteroscopy. Fertility & Sterility

2011;95(1):264-71.

379

(103) Cooper NA, Khan KS, Clark TJ. Local anaesthesia for pain control during outpatient

hysteroscopy: systematic review and meta-analysis. BMJ 2010;340.

(104) Ahmad G, O'Flynn H, Attarbashi S, Duffy JM, Watson A. Pain relief for outpatient

hysteroscopy. Cochrane Database of Systematic Reviews 2010;(11).

(105) Naim NM, Mahdy ZA, Ahmad S, Razi ZRM. The Vabra aspirator versus the Pipelle

device for outpatient endometrial sampling. Australian and New Zealand Journal of

Obstetrics and Gynaecology 2007;47(2):132-6.

(106) Selk A, Kroft J. Misoprostol in operative hysteroscopy: a systematic review and

meta-analysis. Obstetrics & Gynecology 2011;118(4):941-9.

(107) Gkrozou F, Koliopoulos G, Vrekoussis T, Valasoulis G, Lavasidis L, Navrozoglou I,

et al. A systematic review and meta-analysis of randomized studies comparing

misoprostol versus placebo for cervical ripening prior to hysteroscopy. European


(108) Gupta JK, Clark TJ, More S, Pattison H. Patient anxiety and experiences associated

with an outpatient "one-stop" "see and treat" hysteroscopy clinic. Surgical

Endoscopy 2004;18(7):1099-104.

(109) Panel P, Grosdemouge I. Predictive factors of Essure implant placement failure:

prospective, multicenter study of 495 patients. Fertility & Sterility 2010;93(1):29-

34.

(110) Madari S, Al-Shabibi N, Papalampros P, Papadimitriou A, Magos A. A randomised

trial comparing the H Pipelle with the standard Pipelle for endometrial sampling at

'no-touch' (vaginoscopic) hysteroscopy. BJOG 2009;(116):32-7.

(111) Clark TJ, Mahajan D, Sunder P, Gupta JK. Hysteroscopic treatment of symptomatic

submucous fibroids using a bipolar intrauterine system: a feasibility study. European


(112) Zikopoulos KA, Kolibianakis EM, Tournaye H, de ML, Devroey P, Camus M.

Hysteroscopic septum resection using the Versapoint system in subfertile women.

Reproductive BioMedicine Online 2003;7(3):365-7.

(113) Sui L, Wan Q, Zheng RL, Chen M, Xie F, Diao WJ, et al. Transcervical incision of

septa: 447 cases. Surgical Endoscopy 2009;23(9):2078-84.

(114) Zikopoulos KA, Kolibianakis EM, Platteau P, de ML, Tournaye H, Devroey P, et al.

Live delivery rates in subfertile women with Asherman's syndrome after

hysteroscopic adhesiolysis using the resectoscope or the Versapoint system.

Reproductive BioMedicine Online 2004;8(6):720-5.

(115) Garuti G, Centinaio G, Luerti M. Outpatient Hysteroscopic Polypectomy in

Postmenopausal Women: A Comparison between Mechanical and Electrosurgical

Resection. Journal of Minimally Invasive Gynecology 2008;15(5):595-600.

380

(116) Jacobs VR, Paepke S, Schwarz-Boeger U, Fischer T, Pildner Von SS, Plattner B, et

al. Development of a thinner and more flexible type of minihysteroscope with a

controlled 90-degree bendable tip for vision-guided endometrium biopsy. Journal of


(117) Clark TJ, Godwin J, Khan KS, Gupta JK. Ambulatory endoscopic treatment of

symptomatic benign endometrial polyps. A feasibility study. Gynaecological

Endoscopy 2002;11(2-3):91-7.

(118) Sinha D, Kalathy V, Gupta JK, Clark TJ. The feasibility, success and patient

satisfaction associated with outpatient hysteroscopic sterilisation. BJOG: An

International Journal of Obstetrics & Gynaecology 2007;114(6):676-83.

(119) Kelly AJ. Development of RCOG Green-top Guidelines: Policies and Processes

(Clinical Governance Advice 1a). 2013. The Royal College of Obstetricians and

Gynaecologists. http://www.rcog.org.uk/green-top-development Last accessed

25/05/2013. Online Source

(120) Scottish Intercollegiate Guidelines Network. SIGN 50 A guideline developer's

handbook. [Second]. 2011. http://www.sign.ac.uk/guidelines/fulltext/50/ Last

accessed 25/05/2013. Online Source

(121) Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Schunn., et al. GRADE:

What is "quality of evidence" and why is it important to clinicians? BMJ 2008 May

3;336:995-8.

(122) GRADE working group. Grading quality of evidence and strength of

recommendations. BMJ 2004;328:1490.

(123) GRADE working group. http://www.gradeworkinggroup.org/ . 2013Last accessed


(124) lark TJ, Cooper NAM, Kremer C. Best Practice in Outpatient Hysteroscopy. 2011.

RCOG/BSGE. http://www.rcog.org.uk/womens-health/clinical-

guidance/hysteroscopy-best-practice-outpatient-green-top-59 Last accessed


(125) Harbour R, Miller J. A new system for grading recommendations in evidence based

guidelines. BMJ 2001;323(7308):334-6.

(126) GRADEprofiler (GRADEpro) [computer program]. Version 3.6 2011.

(127) Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al.

GRADE: An emerging consensus on rating quality of evidence and strength of

recommendations. BMJ 2008 Apr 26;336:924-6.

(128) Ibargoyen-Roteta N, Gutierrez-Ibarluzea I, Rico-Iturrioz R, Lopez-Argumedo M,

Reviriego-Rodrigo E, Cabriada-Nuno JL, et al. The GRADE approach for assessing

381

new technologies as applied to apheresis devices in ulcerative colitis.

Implementation Science 2010;5.

(129) De R, Liberati A, Ciccone G, Bandieri E, Belfiglio M, Ceccarelli M, et al.

Developing clinical recommendations for breast, colorectal, and lung cancer

adjuvant treatments using the GRADE system: a study from the Programma Ricerca

e Innovazione Emilia Romagna Oncology Research Group. Journal of Clinical

Oncology 2008;26(7):1033-9.

(130) Cuello CA, Pacheco KP, Perez G. Grading recommendations in clinical practice

guidelines: randomised experimental evaluation of four different systems. Archives

of Disease in Childhood 2011;96(8):723-8.

(131) Calderon C, Rotaeche R, Etxebarria A, Marzo M, Rico R, Barandiaran M. Gaining

insight into the Clinical Practice Guideline development processes: qualitative study

in a workshop to implement the GRADE proposal in Spain. BMC Health Services

Research 2006;6(138).

(132) Fox CE, Tirlapur SA, Gulmezoglu AM, Souza JP, Khan KS, EBM-CONNECT

EBMC. Assimilating evidence quality at a glance using graphic display: research

synthesis on labor induction. Acta Obstetricia et Gynecologica Scandinavica

2012;91(8):885-92.

(133) Bauer DT, Guerlain S, Brown PJ. The design and evaluation of a graphical display

for laboratory data. Journal of the American Medical Informatics Association

2010;17(4):416-24.

(134) Khan KS, Borowiack E, Roos C, Kowalska M, Zapalska A, Mol BW, et al. Making

GRADE accessible: a proposal for graphic display of evidence quality assessments.

Evidence Based Medicine 2011;16(3):65-9.

(135) Coulter A, Noone A, Goldacre M. General practitioners' referrals to specialist

outpatient clinics. I. Why general practitioners refer patients to specialist outpatient

clinics. BMJ 1989 Jul 29;299:304.

(136) Bhattacharya S, Middleton LJ, Tsourapas A, Lee AJ, Champaneria R, Daniels JP, et

al. Hysterectomy, endometrial ablation and mirena for heavy menstrual bleeding: A

systematic review of clinical effectiveness and cost-effectiveness analysis. Health

Technology Assessment 2011;15(19):1-252.

(137) Coulter A, McPherson K, Vessey M. Do British women undergo too many or too

few hysterectomies? Social Science & Medicine 1988;27(9):987-94.

(138) Coulter A, kelland J, Long A, Melvill A, O'Meara S. The management of

menorrhagia. Effective Health Care 1995;9.

(139) Coulter A, Peto V, Jenkinson C. Quality of life and patient satisfaction following

treatment for menorrhagia. Family Practice 1994;11(4):394-401.

382

(140) Clark TJ, Khan KS, Foon R, Pattison H, Bryan S, Gupta JK. Quality of life

instruments in studies of menorrhagia: A systematic review. European Journal of

Obstetrics Gynecology and Reproductive Biology 2002;104(2):96-104.

(141) Jones GL, Kennedy SH, Jenkinson C. Health-related quality of life measurement in

women with common benign gynecologic conditions: A systematic review.

American Journal of Obstetrics and Gynecology 2002;187(2):501-11.

(142) National Institute for Clinical Excellence. CG44 Heavy menstrual bleeding: NICE

guideline 2007. http://www.nice.org.uk/CG44 Last accessed 25/05/2013 Online

Source

(143) Royal College of Obstetricians and Gynaecologists. The initial management of

menorrhagia. Evidence based guidelines No. 1. London: Royal College of

Obstetricians and Gynaecologists; 1998.

(144) Reid PC, Coker A, Coltart R. Assessment of menstrual blood loss using a pictorial

chart: A validation study. British Journal of Obstetrics and Gynaecology

2000;107(3):320-2.

(145) Wyatt KM, Dimmock PW, Walker TJ, O'Brien PM. Determination of total

menstrual blood loss. Fertility and Sterility 2001;76(1):125-31.

(146) Chimbira TH, Anderson AB, Naish C, Cope E, Turnbull AC. Reduction of

menstrual blood loss by danazol in unexplained menorrhagia: lack of effect of

placebo. British Journal of Obstetrics & Gynaecology 1980;87(12):1152-8.

(147) Fraser IS, McCarron G, Markham R. A preliminary study of factors influencing

perception of menstrual blood loss volume. American Journal of Obstetrics &

Gynecology 1984;149(7):788-93.

(148) Hallberg L, Nilsson L. Determination of menstrual blood loss. Scandinavian Journal

of Clinical & Laboratory Investigation 1964;16:244-8.

(149) Kurman RJ, Kaminski PF, Norris HJ. The behavior of endometrial hyperplasia. A

long-term study of 'untreated' hyperplasia in 170 patients. Cancer 1985;56(2):403-

12.

(150) Clark TJ, Mann CH, Shah N, Khan KS, Song F, Gupta JK. Accuracy of outpatient

endometrial biopsy in the diagnosis of endometrial hyperplasia. Acta Obstetricia et

Gynecologica Scandinavica 2001;80(9):784-93.

(151) Clark TJ, Mann CH, Shah N, Khan KS, Song F, Gupta JK. Accuracy of outpatient

endometrial biopsy in the diagnosis of endometrial cancer: a systematic quantitative

review. BJOG : an international journal of obstetrics and gynaecology

2002;109(3):313-21.

(152) Goldstein SR. Saline infusion sonohysterography. Clinical Obstetrics and

Gynecology 1996;39(1):248-58.

383

(153) Wolman I, Jaffa AJ, Kupferminz M, Geva E, Gordon D, Lessing JB, et al.

Transvaginal sonohysterography: A method for the evaluation of the endometrial

cavity - 4 years experience. Ultrasound International 1999;5(2):79-86.

(154) Ossola MW, Bertulessi C, Iasi L, Bianchini B, Hanozet F, Grassini E, et al.

Comparison of saline infusion sonography to transvaginal echography and

hysteroscopy in the diagnostic evaluation of abnormal uterine bleeding. Italian

Journal of Gynaecology and Obstetrics 1999;11(4):145-50.

(155) Critchley HOD, Warner P, Lee AJ, Brechin S, Guise J, Graham B. Evaluation of

abnormal uterine bleeding: Comparison of three outpatient procedures within

cohorts defined by age and menopausal status. Health Technology Assessment

2004;8(34):iii-77.

(156) Dijkhuizen FP, Mol BW, Bongers MY, lmann HA, Heintz AP. Cost-effectiveness of

transvaginal sonography and saline infused sonography in the evaluation of

menorrhagia. International Journal of Gynaecology & Obstetrics 2003;83(1):45-52.

(157) Milsom I, Andersson K, Andersch B, Rybo G. A comparison of flurbiprofen,

tranexamic acid, and a levonorgestrel-releasing intrauterine contraceptive device in

the treatment of idiopathic menorrhagia. American Journal of Obstetrics and

Gynecology 1991;164(3):879-83.

(158) Lahteenmaki P, Haukkamaa M, Puolakka J, Riikonen U, Sainio S, Suvisaari J, et al.

Open randomised study of use of levonorgestrel releasing intrauterine system as

alternative to hysterectomy. British Medical Journal 1998;316(7138):1122-6.

(159) Hurskainen R, Teperi J, Rissanen P, Aalto AM, Grenman S, Kivela A, et al. Clinical

outcomes and costs with the levonorgestrel-releasing intrauterine system or

hysterectomy for treatment of menorrhagia: randomized trial 5-year follow-up.

JAMA 2004;291(12):1456-63.

(160) Hurskainen R, Teperi J, Rissanen P, Aalto AM, Grenman S, Kivela A, et al. Quality

of life and cost-effectiveness of levonorgestrel-releasing intrauterine system versus

hysterectomy for treatment of menorrhagia: A randomised trial. Lancet

2001;357(9252):273-7.

(161) Hurskainen R, Teperi J, Aalto AM, Grenman S, Kivela A, Kujansuu E, et al.

Levonorgestrel-releasing intrauterine system or hysterectomy in the treatment of

essential menorrhagia: predictors of outcome. Acta Obstetricia et Gynecologica

Scandinavica 2004;83(4):401-3.

(162) Varma R, Soneja H, Bhatia K, Ganesan R., Rollason T, Clark TJ, et al. The

effectiveness of a levonorgestrel-releasing intrauterine system (LNG-IUS) in the

treatment of endometrial hyperplasia- A long-term follow-up study. European

Journal of Obstetrics Gynecology and Reproductive Biology 2008;139:169-75.

384

(163) Gallos ID, Krishan P, Shehmar M, Gupta JK. LNG-IUS vs. oral progestogens for

endometrial hyperplasia. A comparative cohort study with long term follow-up.

Gynecological Surgery 2009;6.

(164) Cooper KG, Jack SA, Parkin DE, Grant AM. Five-year follow up of women

randomised to medical management or transcervical resection of the endometrium

for heavy menstrual loss: clinical and quality of life outcomes. BJOG: An

International Journal of Obstetrics & Gynaecology 2001;108(12):1222-8.

(165) Lethaby A, Shepperd S, Farquhar C, Cooke I. Endometrial resection and ablation

versus hysterectomy for heavy menstrual bleeding. Cochrane Database of

Systematic Reviews . 2009. Online Source

(166) Nathani F, Clark TJ. Uterine polypectomy in the management of abnormal uterine

bleeding: A systematic review. Journal of Minimally Invasive Gynecology

2006;13(4):260-8.

(167) Lieng M, Istre O, Qvigstad E. Treatment of endometrial polyps: a systematic review.

Acta Obstetricia et Gynecologica Scandinavica 2010;89(8):992-1002.

(168) Hart R, Molnar V, Magos A. Long-term follow-up of hysteroscopic myomectomy

assessed by survival analysis. BJOG : an international journal of obstetrics and

gynaecology 1999;106(7):700-5.

(169) Emanuel MH, Wamsteker K, Hart AA, Metz G, Lammes FB. Long-term results of

hysteroscopic myomectomy for abnormal uterine bleeding. Obstetrics &

Gynecology 1999;93(5 Pt 1):743-8.

(170) Vercellini P, Zaina B, Yaylayan L, Pisacreta A, De O, Crosignani PG.

Hysteroscopic myomectomy: long-term effects on menstrual pattern and fertility.

Obstetrics & Gynecology 1999;94(3):341-7.

(171) Varma R, Soneja H, Clark TJ, Gupta JK. Hysteroscopic myomectomy for

menorrhagia using Versascope bipolar system: efficacy and prognostic factors at a

minimum of one year follow up. European Journal of Obstetrics, Gynecology, &

Reproductive Biology 2009;142(2):154-9.

(172) Clark TJ, Khan KS, Gupta JK. Current practice for the treatment of benign

intrauterine polyps: a national questionnaire survey of consultant gynaecologists in

UK. European Journal of Obstetrics, Gynecology, & Reproductive Biology

2002;103(1):65-7.

(173) Litta P, Cosmi E, Saccardi C, Esposito C, Rui R, Ambrosini G. Outpatient operative

polypectomy using a 5 mm-hysteroscope without anaesthesia and/or analgesia:

Advantages and limits. European Journal of Obstetrics Gynecology and

Reproductive Biology 2008;139(2):210-4.

385

(174) Capobianco G, Vargiu N, Dessole F, Milia L, Dessole S. Office hysteroscopy for

uterine intracavitary pathologies: See and treat approach. Gynecological Surgery

2009;6.

(175) Marsh FA, Rogerson LJ, Duffy SRG. A randomised controlled trial comparing

outpatient versus daycase endometrial polypectomy. BJOG: An International Journal

of Obstetrics and Gynaecology 2006;113(8):896-901.

(176) OPT trial management group. OPT: The Outpatient Polyp Treatment Trial. Protocol.

2010.

http://www.download.bham.ac.uk/bctu/OPT/Trial%20Documentation_Consent%20

PIS%20and%20GP%20Sheets/OPT%20Protocol%20Version%201.3%2015%20Feb

%202008%20FINALdoc.pdf Last accessed 25/05/2013. Online Source

(177) Pelage JP, Le O, Soyer P, Kardache M, Dahan H, Abitbol M, et al. Fibroid-related

menorrhagia: treatment with superselective embolization of the uterine arteries and

midterm follow-up. Radiology 2000;215(2):428-31.

(178) Lefebvre G, Vilos G, Allaire C, Jeffrey J, Arneja J, Birch C, et al. The management

of uterine leiomyomas. Journal of Obstetrics & Gynaecology Canada: JOGC

2003;25(5):396-418.

(179) Ware JE, Jr., Sherbourne CD. The MOS 36-item short-form health survey (SF-36).

I. Conceptual framework and item selection. Medical Care 1992;30(6):473-83.

(180) Jenkinson C, Peto V, Coulter A. Making sense of ambiguity: evaluation in internal

reliabilityand face validity of the SF 36 questionnaire in women presenting with

menorrhagia. Quality in Health Care 1996;5:9-12.

(181) Lamping DL, Rowe P, Clarke A, Black N, Lessof L. Development and validation of

the Menorrhagia Outcomes Questionnaire. British Journal of Obstetrics &

Gynaecology 1998;105(7):766-79.

(182) Shaw RW, Brickley MR, Evans L, Edwards MJ. Perceptions of women on the

impact of menorrhagia on their health using multi-attribute utility assessment.

British Journal of Obstetrics & Gynaecology 1998;105(11):1155-9.

(183) Crow R, Gage H, Hampson S, Hart J, Kimber A, Storey L, et al. The measurement

of satisfaction with healthcare: implications for practice from a systematic review of

the literature. Health Technology Assessment (Winchester, England) 2002;6(32):1-

244.

(184) Middleton LJ, Champaneria R, Daniels JP, Bhattacharya S, Cooper KG, Hilken NH,

et al. Hysterectomy, endometrial destruction, and levonorgestrel releasing

intrauterine system (Mirena) for heavy menstrual bleeding: Systematic review and

meta-analysis of data from individual patients. BMJ 2010;341(7769).

(185) TreeAge Pro 2009 Suite [computer program]. Version 1.0.2 2009.

386

(186) Higham JM, Shaw RW. Clinical associations with objective menstrual blood

volume. European Journal of Obstetrics Gynecology and Reproductive Biology

1999;82:73-6.

(187) Fuentes JA. In-office hysteroscopic polypectomy: Eight years analysis [Spanish]

Polipectomia histeroscopica en consultorio: Analisis de ocho anos. Ginecologia y

Obstetricia de Mexico 2007;75(6):341-6.

(188) Gonzlez A, Fernndez J, Rodrguez A. Operative office hysteroscopy. Gynecological

Surgery 2010;7.

(189) Bayer Healthcare. Mirena® Prescribing Information. 2009

http://labeling.bayerhealthcare.com/html/products/pi/Mirena_PI.pdf. Last accessed


(190) Hologic Inc. NovaSure® Instructions for Use. 2009

http://www.novasure.com/pdf/info/Instructions.pdf Last accessed 25/02013. Online

Source

(191) Ethicon Women's Health and Urology. Instructions for use gynecare®

thermachoice® iii uterine balloon therapy system. 2010.

http://www.jjnordic.com/Lists/FileList1/Attachments/334/Thermachoice%20IFU.pd

f Last accessed 25/05/2013. Online Source

(192) Kaunitz AM, Bissonnette F, Monteiro I, Lukkari-Lax E, Muysers C, Jensen JT.

Levonorgestrel-releasing intrauterine system or medroxyprogesterone for heavy

menstrual bleeding: a randomized controlled trial. Obstetrics & Gynecology

2010;116(3):625-32.

(193) Lethaby A, Hickey M, Garry R, Penninx J. Endometrial resection/ablation

techniques for heavy menstrual bleeding. Cochrane Database of Systematic Reviews

2009;(4).

(194) Lethaby A, Shepperd S, Cooke I, Farquhar C. Endometrial resection and ablation

versus hysterectomy for heavy menstrual bleeding. Cochrane Database of

Systematic Reviews 2000;(2).

(195) The REST Investigators. Uterine Artery Embolization versus surgery for

symptomatic uterine fibroids. New England Journal of Medicine 2007;(536):360-70.

(196) FIGO Committee on Gynecologic Oncology. Staging Classifications and Clinical

Practice Guidelines for Gynaecological Cancers. 2000.

http://www.ginecologia.unipd.it/Assistenza-

Documenti/Unita'%20operative/Ginecologia%20Oncologica/staging_booklet.pdf

Last accessed 25/05/2013 Online Source

(197) de Kroon CD, de Bock GH, Dieben SW, Jansen FW. Saline contrast

hysterosonography in abnormal uterine bleeding: A systematic review and meta-

387

analysis. BJOG: An International Journal of Obstetrics and Gynaecology

2003;110(10):938-47.

(198) Clark TJ, Voit D, Gupta JK, Hyde C, Song F, Khan KS. Accuracy of hysteroscopy

in the diagnosis of endometrial cancer and hyperplasia: a systematic quantitative

review. JAMA 2002;288(13):1610-21.

(199) Farquhar C, Ekeroma A, Furness S, Arroll B. A systematic review of transvaginal

ultrasonography, sonohysterography and hysteroscopy for the investigation of

abnormal uterine bleeding in premenopausal women. Acta Obstetricia et

Gynecologica Scandinavica 2003;82(6):493-504.

(200) Giorda G, Crivellari D, Veronesi A, Perin T, Campagnutta E, Carbone A, et al.

Comparison of ultrasonography, hysteroscopy, and biopsy in the diagnosis of

endometrial lesions in postmenopausal tamoxifen-treated patients. Acta Obstetricia

et Gynecologica Scandinavica 2002;81(10):975-80.

(201) Mara M, Maskova J, Fucikova Z, Kuzel D, Belsan T, Sosna O. Midterm clinical and

first reproductive results of a randomized controlled trial comparing uterine fibroid

embolization and myomectomy. Cardiovascular & Interventional Radiology

2008;31(1):73-85.

(202) Emanuel MH, Verdel MJC, Stas H, Wamsteker K, Lammes FB. An audit of true

prevalence of intra-uterine pathology: The hysteroscopical findings controlled for

patient selection in 1202 patients with abnormal uterine bleeding. Gynaecological

Endoscopy 1995;4(4):237-41.

(203) Dreisler E, Stampe SS, Ibsen PH, Lose G. Prevalence of endometrial polyps and

abnormal uterine bleeding in a Danish population aged 20-74 years. Ultrasound in

Obstetrics and Gynecology 2009;33(1):102-8.

(204) Lasmar RB, Dias R, Barrozo PR, Oliveira MA, Coutinho ES, da Rosa DB.

Prevalence of hysteroscopic findings and histologic diagnoses in patients with

abnormal uterine bleeding. Fertility & Sterility 2008;89(6):1803-7.

(205) van Dongen H, de Kroon CD, Jacobi CE, Trimbos JB, Jansen FW. Diagnostic

hysteroscopy in abnormal uterine bleeding: a systematic review and meta-analysis.

BJOG: An International Journal of Obstetrics & Gynaecology 2007;114(6):664-75.

(206) Clevenger-Hoeft M, Syrop CH, Stovall DW, Van Voorhis BJ. Sonohysterography in

premenopausal women with and without abnormal bleeding. Obstetrics &

Gynecology 1999;94(4):516-20.

(207) Lurie S, Piper I, Woliovitch I, Glezerman M. Age-related prevalence of

sonographicaly confirmed uterine myomas. Journal of Obstetrics and Gynaecology

2005;25(1):42-3.

(208) Iram S, Musonda P, Ewies AAA. Premenopausal bleeding: When should the

endometrium be investigated?-A retrospective non-comparative study of 3006

388

women. European Journal of Obstetrics Gynecology and Reproductive Biology

2010;148(1):86-9.

(209) Basu A, Lewis P. Transvaginal scan, saline infusion sonography and hysteroscopy in

abnormal uterine bleeding. International Journal of Gynecology and Obstetrics

2009;107:S120-S121.

(210) Smith-Bindman R. Diagnostic imaging in the differential diagnosis of vaginal

bleeding and breast mass. Advanced Studies in Medicine 2004;4(9):476-82.

(211) Grimbizis GF, Tsolakidis D, Mikos T, Anagnostou E, Asimakopoulos E,

Stamatopoulos P, et al. A prospective comparison of transvaginal ultrasound, saline

infusion sonohysterography, and diagnostic hysteroscopy in the evaluation of

endometrial pathology. Fertility and Sterility 2010;94(7):2720-5.

(212) Botsis D, Kassanos D, Antoniou G, Pyrgiotis E, Karakitsos P, Kalogirou D.

Adenomyoma and leiomyoma: Differential diagnosis with transvaginal sonography.

Journal of Clinical Ultrasound 1997;26(1):21-5.

(213) Vercellini P, Cortesi I, Oldani S, Moschetta M, De O, Crosignani PG. The role of

transvaginal ultrasonography and outpatient diagnostic hysteroscopy in the

evaluation of patients with menorrhagia. Human Reproduction 1997;12(8):1768-71.

(214) Aslam M, Ijaz L, Tariq S, Shafqat K, Meher UN, Ashraf R, et al. Comparison of

transvaginal sonography and saline contrast sonohysterography in women with

abnormal uterine bleeding: correlation with hysteroscopy and histopathology.

International Journal of Health Sciences 2007;1(1):17-24.

(215) Goldschmit R, Katz Z, Blickstein I, Caspi B, Dgani R. The accuracy of endometrial

pipelle sampling with and without sonographic measurement of endometrial

thickness. Obstetrics and Gynecology 1993;82(5):727-30.

(216) Evidence Based Medicine T. Stats Calculator.

http://ktclearinghouse.ca/cebm/practise/ca/calculators/statscalc 2011. Last accessed


(217) Angioni S, Loddo A, Milano F, Piras B, Minerba L, Melis GB. Detection of Benign

Intracavitary Lesions in Postmenopausal Women with Abnormal Uterine Bleeding:

A Prospective Comparative Study on Outpatient Hysteroscopy and Blind Biopsy.


(218) Lethaby AE, Cooke I, Rees M. Progesterone or progestogen-releasing intrauterine

systems for heavy menstrual bleeding. Cochrane Database of Systematic Reviews

2005;(4).

(219) Longinotti MK, Jacobson GF, Hung YY, Learman LA. Probability of hysterectomy

after endometrial ablation. Obstetrics and Gynecology 2008;112(6):1214-20.

389

(220) Hirst A, Dutton S, Wu O, Briggs A, Edwards C, Waldenmaier L, et al. A multi-

centre retrospective cohort study comparing the efficacy, safety and cost-

effectiveness of hysterectomy and uterine artery embolisation for the treatment of

symptomatic uterine fibroids. The HOPEFUL study. Health Technology Assessment

(Winchester, England) 2008;12(5):1-248, iii.

(221) Pinto I, Chimeno P, Romo L, Haya J, De la Cal M, Bajo J. Uterine fibroids: Uterine

artery embolization versus abdominal hysterectomy for treatment. A prospective

randomized and controlled trial. Radiology 2003;226(2):425-31.

(222) van Dongen H, Janssen CA, Smeets MJ, Emanuel MH, Jansen FW. The clinical

relevance of hysteroscopic polypectomy in premenopausal women with abnormal

uterine bleeding. BJOG: An International Journal of Obstetrics & Gynaecology

2009;116(10):1387-90.

(223) Lieng M, Istre O, Sandvik L, Engh V, Qvigstad E. Clinical effectiveness of

transcervical polyp resection in women with endometrial polyps: randomized

controlled trial. Journal of Minimally Invasive Gynecology 2010;17(3):351-7.

(224) Camanni M, Bonino L, Delpiano EM, Ferrero B, Migliaretti G, Deltetto F.

Hysteroscopic management of large symptomatic submucous uterine myomas.


(225) Bourdel N, Bonnefoy C, Jardon K, Da D, Tognazza E, Rabischong B, et al.

[Hysteroscopic myomectomy: recurrence and satisfaction survey at short- and long-

term]. [French] Myomectomie hysteroscopique: recidive et enquete de satisfaction a

court et long terme. Journal de Gynecologie, Obstetrique et Biologie de la

Reproduction 2011;40(2):116-22.

(226) Cravello L, Farnarier J, Roger V, D'Ercole C, Blanc B. Hysteroscopic myomectomy.

Functional results with an average follow-up of 6 years [French] Myomectomie

hysteroscopique. Resultats fonctionnels avec un recul moyen de 6 ans. Journal de

Gynecologie, Obstetrique et Biologie de la Reproduction 1998;27(6):593-7.

(227) Munoz JL, Jimenez JS, Hernandez C, Vaquero G, Perez C, Noguero R, et al.

Hysteroscopic myomectomy: our experience and review. Journal of the Society of

Laparoendoscopic Surgeons 2003;7(1):39-48.

(228) Polena V, Mergui JL, Perrot N, Poncelet C, Barranger E, Uzan S. Long-term results

of hysteroscopic myomectomy in 235 patients. European Journal of Obstetrics,

Gynecology, & Reproductive Biology 2007;130(2):232-7.

(229) Haynes PJ, Hodgson H, Anderson AB, Turnbull AC. Measurement of menstrual

blood loss in patients complaining of menorrhagia. British Journal of Obstetrics &

Gynaecology 1977;84(10):763-8.

(230) van der Kooji SM, Bipat S, Hehenkamp WJ, Ankum WM, Reekers JA. Uterine

artery embolization versus surgery in the treatment of symptomatic fibroids: a

390

systematic review and metaanalysis. American Journal of Obstetrics & Gynecology

2011;205(4):317.

(231) Gupta JK, Sinha AS, Lumsden MA, Hickey M. Uterine artery embolization for

symptomatic uterine fibroids. Cochrane database of systematic reviews (Online)

2006;(1).

(232) Personal Social Services Research Unit. Unit Costs of Health and Social Care 2010.

http://www.pssru.ac.uk/project-pages/unit-costs/2010/index.php Last accessed

25/05/2013 Online Source

(233) Clark TJ, Samuel N, Malick S, Middleton LJ, Daniels J, Gupta JK. Bipolar

Radiofrequency Compared With Thermal Balloon Endometrial Ablation in the

Office: A Randomized Controlled Trial. Obstetrics & Gynecology 2011;117(1).

(234) Sculpher M. A cost-utility analysis of abdominal hysterectomy versus transcervical

endometrial resection for the surgical treatment of menorrhagia. International

Journal of Technology Assessment in Health Care 1998;14(2):302-19.

(235) Department of Health. NHS reference costs 2009-2010.

http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyA

ndGuidance/DH_123459 2011.Last accessed 25/05/2013 Online Source

(236) Reinhold C, McCarthy S, Bret PM, Mehio A, Atri M, Zakarian R, et al. Diffuse

adenomyosis: Comparison of endovaginal US and MR imaging with histopathologic

correlation. Radiology 1996;199(1):151-8.

(237) Timmermans A, van H, Mol BW, Veersema S, Jansen FW. Hysteroscopy and

removal of endometrial polyps: a Dutch survey. European Journal of Obstetrics,

Gynecology, & Reproductive Biology 2008;138(1):76-9.

(238) Brechin ST, Cameron AM, Paterson AR, Williams HOD, Critchley. Intrauterine

polyps - A cause of unscheduled bleeding in women using the levonorgestrel

intrauterine system. Human Reproduction 2000;15(3):650-2.

(239) Mercorio F, De R, Di A, Cerrota G, Bifulco G, Vanacore F, et al. The effect of a

levonorgestrel-releasing intrauterine device in the treatment of myoma-related

menorrhagia. Contraception 2003;67(4):277-80.

(240) Timmermans A, Veersema S, Van TC, Van LF, V, Opmeer BC, Bongers MY, et al.

Should endometrial polyps be removed in patients with postmenopausal bleeding? -

An assessment of study designs and report of a failed randomised controlled trial

(ISRCTN73825127). BJOG: An International Journal of Obstetrics and

Gynaecology 2009;116(10):1391-5.

(241) Di Spiezo Sardo A, Bettocchi S, Spinelli M, Guida M, Nappi L, Angioni S, et al.

Review of New Office-Based Hysteroscopic Procedures 2003-2009. Journal of


Ambulatory gynaecology: guidelines and economic analysis

Documents